U.S. patent application number 12/822834 was filed with the patent office on 2010-12-30 for image compressing apparatus, image compressing method, image decompressing apparatus, image decompressing method, image forming apparatus and recording meduim.
This patent application is currently assigned to SHARP KABUSHIKI KAISHA. Invention is credited to Hideyoshi YOSHIMURA.
Application Number | 20100329548 12/822834 |
Document ID | / |
Family ID | 43380794 |
Filed Date | 2010-12-30 |
![](/patent/app/20100329548/US20100329548A1-20101230-D00000.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00001.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00002.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00003.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00004.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00005.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00006.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00007.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00008.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00009.png)
![](/patent/app/20100329548/US20100329548A1-20101230-D00010.png)
View All Diagrams
United States Patent
Application |
20100329548 |
Kind Code |
A1 |
YOSHIMURA; Hideyoshi |
December 30, 2010 |
IMAGE COMPRESSING APPARATUS, IMAGE COMPRESSING METHOD, IMAGE
DECOMPRESSING APPARATUS, IMAGE DECOMPRESSING METHOD, IMAGE FORMING
APPARATUS AND RECORDING MEDUIM
Abstract
When compressing continuous tone bit map image data, the image
compression section of the image compressing apparatus segments the
continuous tone bit map image data into bit map image data for
lossy compression, index image data for lossless compression and
bit map image data for lossless compression based on pixel
identification information data. The lossy compression section of
the image compression section performs lossy compression process
according to the JPEG method for the bit map image data for lossy
compression, and the first lossless compression section thereof
performs lossless compression process according to the JBIG method
for the index image data for lossless compression. Furthermore, the
second lossless compression section thereof performs lossless
compression process according to the JPEG-LS method for the bit map
image data for lossless compression.
Inventors: |
YOSHIMURA; Hideyoshi;
(Osaka-shi, JP) |
Correspondence
Address: |
BIRCH STEWART KOLASCH & BIRCH
PO BOX 747
FALLS CHURCH
VA
22040-0747
US
|
Assignee: |
SHARP KABUSHIKI KAISHA
Osaka-shi
JP
|
Family ID: |
43380794 |
Appl. No.: |
12/822834 |
Filed: |
June 24, 2010 |
Current U.S.
Class: |
382/164 ;
382/244 |
Current CPC
Class: |
H04N 19/103 20141101;
H04N 19/12 20141101; H04N 1/00278 20130101; H04N 19/17 20141101;
H04N 19/186 20141101; H04N 1/413 20130101; H04N 19/136
20141101 |
Class at
Publication: |
382/164 ;
382/244 |
International
Class: |
G06K 9/34 20060101
G06K009/34; G06K 9/36 20060101 G06K009/36 |
Foreign Application Data
Date |
Code |
Application Number |
Jun 25, 2009 |
JP |
2009-151206 |
Claims
1. An image compressing apparatus for compressing an input image
having a plurality of color components, comprising: a first image
segmentation section for segmenting said input image into a image
for lossless compression to be subjected to lossless compression
and a image for lossy compression to be subjected to lossy
compression, based on pixel identification information indicating
that respective pixels constituting said input image belong to
which of a plurality of areas including a text area and a
photograph area; a color determination section for determining
color information that is used when said image for lossless
compression is further segmented; a second image segmentation
section for segmenting said image for lossless compression
segmented by said first image segmentation section into a first
image for lossless compression containing said color information
determined by said color determination section and one or a
plurality of second images for lossless compression excluding said
first image for lossless compression; and an image compression
section for performing respectively different compression process
for each of said image for lossy compression segmented by said
first image segmentation section, said first image for lossless
compression and said second image for lossless compression
segmented by said second image segmentation section.
2. The image compressing apparatus according to claim 1, further
comprising: a frequency distribution generating section for
generating frequency distribution of color information of
respective pixels constituting said image for lossless compression
based on said image for lossless compression, wherein said color
determination section determines the color information indicating a
predetermined number of colors whose frequency of occurrence is
higher, based on the frequency distribution generated by said
frequency distribution generating section.
3. The image compressing apparatus according to claim 1, wherein
said color determination section determines the color information
for respective predetermined areas in said input image.
4. The image compressing apparatus according to claim 1, wherein
said first image segmentation section segments said input image
into said image for lossless compression and said image for lossy
compression according to a predetermined condition.
5. The image compressing apparatus according to claim 1, wherein
said second image segmentation section generates said first image
for lossless compression by replacing a pixel value of a pixel
included in a pixel constituting said image for lossless
compression segmented by said first image segmentation section and
segmented into said second image for lossless compression with a
pixel value indicating a transparent color.
6. An image compressing method for compressing an input image
having a plurality of color components, comprising: a first image
segmentation step for segmenting said input image into a image for
lossless compression to be subjected to lossless compression and a
image for lossy compression to be subjected to lossy compression,
based on pixel identification information indicating that
respective pixels constituting said input image belong to which of
a plurality of areas including a text area and a photograph area; a
color determination step for determining color information that is
used when said image for lossless compression is further segmented;
a second image segmentation step for segmenting said image for
lossless compression segmented at said first image segmentation
step into a first image for lossless compression containing said
color information determined at said color determination step and
one or a plurality of second images for lossless compression
excluding said first image for lossless compression; and an image
compression step for performing respectively different compression
process for each of said image for lossy compression segmented by
said first image segmentation step, said first image for lossless
compression and said second image for lossless compression
segmented by said second image segmentation step.
7. An image decompressing apparatus for decompressing a compressed
image, comprising: an image decompression section for performing
respectively different decompression process for a lossy
compression image subjected to lossy compression after segmented
from one image, and a plurality of lossless compression images each
subjected to different types of lossless compression after
segmented from said one image; and an image composing section for
composing a plurality of images obtained by performing the
different types of decompression process by said image
decompression section, by superimposing the plurality of images in
a predetermined sequence.
8. An image decompressing method for decompressing a compressed
image, comprising: an image decompression step for performing
respectively different decompression process for a lossy
compression image subjected to lossy compression after segmented
from one image, and a plurality of lossless compression images each
subjected to different types of lossless compression after
segmented from said one image; and an image composing step for
composing a plurality of images obtained by performing the
different types of decompression process at the image decompression
step, by superimposing the plurality of images in a predetermined
sequence.
9. An image forming apparatus comprising: said image decompressing
apparatus according to claim 7; and an image forming section for
forming an output image based on an image processed by said image
decompressing apparatus.
10. An image forming apparatus comprising: said image compressing
apparatus according to claim 1; said image decompressing apparatus
including an image decompression section for performing
respectively different decompression process for a lossy
compression image subjected to lossy compression after segmented
from one image, and a plurality of lossless compression images each
subjected to different types of lossless compression after
segmented from said one image; and an image composing section for
composing a plurality of images obtained by performing the
different types of decompression process by said image
decompression section, by superimposing the plurality of images in
a predetermined sequence; and an image forming section for forming
an output image based on the image processed by said image
decompressing apparatus, wherein said image decompressing apparatus
performs decompression process for an image compressed by said
image compressing apparatus.
11. A recording medium on which a computer program for causing a
computer to perform compression process for an input image having a
plurality of color components are recorded so as to be readable by
said computer, said computer programs comprising: a first image
segmentation step for causing said computer to segment said input
image into a image for lossless compression to be subjected to
lossless compression and a image for lossy compression to be
subjected to lossy compression, based on pixel identification
information indicating that respective pixels constituting said
input image belong to which of a plurality of areas including a
text area and a photograph area; a color determination step for
causing said computer to determine color information that is used
when said image for lossless compression is further segmented; a
second image segmentation step for causing said computer to segment
said image for lossless compression segmented at said first image
segmentation step into a first image for lossless compression
containing said color information determined at said color
determination step and one or a plurality of second images for
lossless compression excluding said first image for lossless
compression; and an image compression step for causing said
computer to perform respectively different compression process for
each of said image for lossy compression segmented by said first
image segmentation step, said first image for lossless compression
and said second image for lossless compression segmented by said
second image segmentation step.
12. A recording medium on which computer programs for causing a
computer to perform decompression process for a compressed image
are recorded so as to be readable by said computer, said computer
programs comprising: an image decompression step for causing said
computer to perform respectively different decompression process
for a lossy compression image subjected to lossy compression after
segmented from one image, and a plurality of lossless compression
images each subjected to different types of lossless compression
after segmented from said one image; and an image composing step
for causing said computer to compose a plurality of images obtained
by the different types of decompression process at said image
decompression step by superimposing the plurality of images in a
predetermined sequence.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This Nonprovisional application claims priority under 35
U.S.C. .sctn.119(a) on Patent Application No. 2009-151206 filed in
Japan on Jun. 25.2009, the entire contents of which are hereby
incorporated by reference.
BACKGROUND
[0002] 1. Technical Field
[0003] The present invention relates to an image compressing
apparatus and an image compressing method for compressing an image,
an image decompressing apparatus and an image decompressing method
for decompressing a compressed image, an image forming apparatus,
and a recording medium.
[0004] 2. Description of Related Art
[0005] Apparatuses having a printing function, such as those
referred to as a monofunctional printer or a multifunctional
printer, have been widely used. Such a printing apparatus generally
receives image information for forming an output image from an
external personal computer (hereafter referred to as a PC) or the
like via a wired or wireless communication path, forms an output
image by performing conversion process suited for the apparatus
itself and prints the image on paper. Furthermore, generally
speaking, the printing apparatus receives image information from an
external device in a format referred to as PDL (a page description
language) or in a specific format, i.e., in a format represented by
a form close to that of the output image peculiar to the
apparatus.
[0006] The PDL format is a format defined to provide versatility so
that printing can be accomplished by various printers without
causing the user to be aware of the characteristics of the printing
apparatuses as much as possible. As an example of PDL, PostScript
developed by Adobe Systems Incorporated is available for example.
On the other hand, such a specific format is frequently used to
reduce the load on the side of a printing apparatus or on the side
of an image information sender although versatility is reduced, and
various specific formats are available.
[0007] The following will explain process for receiving PDL data
written in PDL and performing printing by a conventional printer.
FIG. 1 is a block diagram showing the internal configuration of the
conventional printer. The conventional printer 501 transmits and
receives information to and from an external PC 500 via a wired or
wireless communication path 500a, such as a USB (universal serial
bus) cable or a wireless LAN (local area network).
[0008] The printer 501 obtains PDL data (image information) from
the external PC 500 via the communication path 500a. The PDL data
input to the printer 501 is subjected to RIP process by an RIP
(raster image processor) process section 502. In the RIP process,
the input PDL data is interpreted according to PDL language
specifications and continuous tone bit map image data is generated.
Since a print output format conversion process section 503 performs
pseudo tone reproduction using density gradation that can be
reproduced by a print engine 504, the continuous tone bit map image
data is subjected to tone reproduction process using image process,
such as a dither method and an error diffusion method and then
converted into halftone bit map image data.
[0009] The halftone bit map image data is input to the print engine
504 equipped with an electrophotographic type or ink-jet type
printing section. The print engine 504 performs printing on paper
according to the input halftone bit map image data. The RIP process
section 502 and the print output format conversion process section
503 are constituted by a processor or an ASIC (application specific
integrated circuit) (both are not shown) mounted in the printer 501
or a combination of these, for example.
[0010] Some type of the printer 501 is configured so that a
plurality of print outputs can be performed for the same PDL data.
The printer 501 capable of performing a plurality of print outputs
stores image data therein in any one of image formats, such as a
format received from the external PC 500, an intermediate format
obtained by converting the received format or a format according to
a final output image format. In addition, the printer 501 performs
a plurality of print outputs using the stored image data, thereby
attaining a function capable of performing a plurality of print
outputs using image data obtained by performing input process
once.
[0011] On storing image data in the format received from the
external PC 500, the amount of process at the time of the reception
is small, but the amount of process at the time of printing is
large. On the other hand, on storing image data in the format
according to the final output image format, the amount of process
at the time of the reception is large, but the amount of process at
the time of printing is small. On storing image data in the
intermediate format, the amount of process is intermediate
therebetween. In order to shorten the time required to perform a
plurality of print outputs, it is efficient to perform the process
of the RIP process section 502 and the print output format
conversion process section 503 only once and to store and reuse the
result in the format according to the final output image
format.
[0012] In addition, on storing image data in the format according
to the intermediate format or the final output image format, a data
storage capacity per page is generally reduced by data compression.
This is based on the facts that the amount of image data to be
printed out is generally large and that when the image data is bit
map data, a relatively large compression rate can be expected.
Hence, it is frequently advantageous that system loads, such as the
data storage capacity and bus bandwidth, increasing when the image
data is processed without compression, can be reduced, even in
consideration of the overhead for data compression and data
decompression.
[0013] FIG. 2 is a block diagram showing the internal configuration
of another conventional printer. Components similar to those shown
in FIG. 1 are designated by the same numerals and their
descriptions are omitted. When the printer 501 shown in FIG. 1
stores image data therein on performing a plurality of print
outputs for the same PDL data, the configuration shown in FIG. 2 in
which halftone bit map image data is compressed and stored is used
efficiently.
[0014] In the printer 501 shown in FIG. 2, a halftone image
compressor 505 compresses the halftone bit map image data that has
been converted from the continuous tone bit map image data by the
print output format conversion process section 503. The compressed
image data is stored in a storage section 506 constituted by a hard
disk or a flash ROM (read only memory). The compressed image data
stored in the storage section 506 is decompressed by a halftone
image decompressor 507, returned to the halftone bit map image data
and input to the print engine 504.
[0015] Since the halftone image compressor 505 and the halftone
image decompressor 507 are used to process halftone images, an
image compressing method being effective in compressing halftone
images is used. An example to be used is JBIG (a binary image
lossless compression method decided by Joint Bi-level Image Experts
Group, ITU-T recommendation T.82)- or JBIG2 (a compression method
for lossless compression or lossy compression for images ranging
from binary images to multilevel images, decided by Joint Bi-level
Image Experts Group 2, ITU-T recommendation T.88), recommended by
Telecommunication Standardization Sector (ITU-T) of ITU
(International Telecommunication Union).
[0016] Furthermore, some type of the printer 501 (a multifunctional
printer) is configured so as to be able to perform print output,
external output, facsimile transmission (hereafter referred to as
FAX transmission), etc. for image data stored inside the printer
501 according to the user's request so that the image data can be
reused. FIG. 3 is a block diagram showing the internal
configuration of still another conventional printer. Components
similar to those shown in FIGS. 1 and 2 are designated by the same
numerals and their descriptions are omitted.
[0017] As shown in FIG. 3, since the printer 501 capable of
performing plural types of process for the stored image data has a
plurality of output destinations to which the stored image data is
output, it is preferable that the image data is stored in an
intermediate format so that the image data can be converted into
various kinds of formats depending on output destination. FIG. 3
shows a configuration in which continuous tone bit map image data
is compressed and stored. In other words, in the printer 501 shown
in FIG. 3, the continuous tone bit map image data generated from
the PDL data by the RIP process section 502 is compressed by a
continuous tone image compressor 508 and then stored in the storage
section 506.
[0018] The compressed image data stored in the storage section 506
is decompressed by a continuous tone image decompressor 509 and
returned to the continuous tone bit map image data and then input
to any one of the print output format conversion process section
503, a JPEG format conversion process section 510 and a FAX format
conversion process section 512. More specifically, when the user
requests printing through the operation section (not shown) of the
printer 501, the continuous tone bit map image data is input to the
print output format conversion process section 503. In addition,
when the user requests image transmission in the JPEG (an image
compression method for lossy compression or lossless compression of
multilevel images, decided by Joint Photographic Experts Group,
ITU-T recommendation T.81) format, the continuous tone bit map
image data is input to the JPEG format conversion process section
510. Furthermore, when the user requests FAX transmission, the
continuous tone bit map image data is input to the FAX format
conversion process section 512.
[0019] The continuous tone bit map image data input to the print
output format conversion process section 503 is converted into
halftone bit map image data by the print output format conversion
process section 503 and printed out onto paper by the print engine
504. The continuous tone bit map image data input to the JPEG
format conversion process section 510 is converted into JPEG image
data by the JPEG format conversion process section 510 and is
transmitted to a destination via a network transmission section 511
capable of performing communication with an external network. The
continuous tone bit map image data input to the FAX format
conversion process section 512 is converted into FAX transmission
format image data by the FAX format conversion process section 512
and is transmitted to a destination via a FAX transmission section
513 that has an interface for FAX and a communication function.
[0020] The JPEG format conversion process section 510 and the FAX
format conversion process section 512 are constituted by a
processor or an ASIC mounted in the printer 501 or a combination of
these, for example.
[0021] The halftone bit map image data for print output, the JPEG
image data for image transmission and the FAX transmission format
image data for FAX transmission are different from each other in
format. Hence, when the respective formats of image data are
created in advance and stored inside the printer 501, the required
process capability and the amount of data to be created increase.
Furthermore, when either format of image data is stored inside the
printer 501 and reused after format conversion, image quality is
degraded.
[0022] Hence, the configuration in which image data of the
continuous tone bit map format serving as an intermediate format
convertible into any format is compressed and stored as shown in
FIG. 3, instead of storing the image data in the respective
formats, is convenient in view of process capability, the amount of
data and data management. However, when the continuous tone image
compressor 508 compresses the continuous tone bit map image data
completely by performing lossless compression process, the rate of
the compression is limited depending on the continuous tone bit map
image data, and only a compression rate limited to some extent is
obtained, although no difference occurs in image quality between
before and after the image compression.
[0023] In recent years, various kinds of methods have been proposed
as image compression methods for multilevel images. Examples of the
methods include lossless compression methods that use
one-dimensional auto-correlation, such as a compression method
based on the run-length method, the LZW method based on the
Lempel-Ziv method serving as a lexicographic compression method,
and the DEFLATE method. Furthermore, the ITU-T Recommendation
stipulates the JPEG lossy compression method based on DCT (discrete
cosine transform), including image compression method information
for reference to the definition of the image decompressing method.
Moreover, Lossless JPEG (ITU-T Recommendation T.81, Annex H)
stipulates a lossless compression method based on two-dimensional
DPCM (differential pulse code modulation). Still further, JPEG-LS
(ITU-T Recommendation T.87) and JPEG 2000 (ITU-T Recommendation
T.800) respectively stipulate a lossless compression method and a
lossy compression method in which methods different from each other
are used. These methods are used widely.
[0024] Some types of apparatuses, such as a digital copier and a
multifunctional printer, have a scanner function for optically
reading a document placed on a document table and generating
digital image data of the read image. Since this type of apparatus
reads the document optically, the digital image data contains noise
and errors caused by a light source and a reading device.
Furthermore, since the optical resolution is limited, edges and the
like cannot be reproduced, and the digital image data generated by
the scanner function cannot reproduce the document completely. For
this reason, when an image generated using the scanner function is
compressed, the lossy compression method defined in JPEG or JPEG
2000 is generally used frequently.
[0025] On the other hand, various kinds of image data, such as text
images of texts, characters and the like, vector images, such as
ruled lines and graphics in graphs, and photograph images in which
image data taken using a digital camera or the like are pasted
partly, are mixed in continuous tone bit map image data that is
generated electronically from PDL data using a method, such as the
RIP process. Among the various kinds of image data, text images and
vector images generally have relatively small numbers of colors and
gray scales. Since the resolution of the visual sense of the human
is high in low gray scales, blurs at edges caused by lossy
compression are easily recognized as degradation in image quality.
Hence, text images and vector images should not be subjected to
lossy compression if possible. Furthermore, since text images and
vector images generally have small numbers of colors, the
compression rates can be improved easily even when lossless
compression is performed.
[0026] Moreover, since photograph images, such as originally
image-sensed images, represented using relatively large numbers of
colors and gray scales have a large number of colors, when such
images are subjected to lossless compression, the compression rates
thereof are not expected to be improved. However, the resolution of
the visual sense of the human is low in multiple gray scales, and
some differences in photograph images cannot be distinguished.
Hence, photograph images are suited for lossy compression that can
further reduce the amount of data although the compression is lossy
compression. For this reason, with respect to continuous tone bit
map image data, it can be expected that the image quality
equivalent to that obtained before the compression can be
maintained by using lossless compression and lossy compression in
combination, while the compression rate is improved.
[0027] For example, a method has been proposed in which a color
image is segmented into plural kinds of portions, such as text
portions and halftone portions or texts, graphics and images, and
the respective portions are subjected to appropriate encoding
process (image compression) (refer to Japanese Patent Application
Laid-open No. H03-104380 (1991) and Japanese Patent Application
Laid-open No. 2000-184205). In addition, an apparatus has been
proposed in which input image data is segmented into areas having
different numbers of gray scales, and one area is encoded using a
reversible encoding method and the other is encoded using
irreversible encoding method (refer to Japanese Patent Application
Laid-open No. 2003-158739).
SUMMARY
[0028] However, when a multifunctional printer segments an image
into a photograph area suited for lossy compression based on the
information written in PDL and performs lossy compression for the
photograph area and performs lossless compression for the remaining
image area into which text and vector images are segmented, there
occurs a problem when image quality is balanced with compression
rate. For example, the image compressor and the image decompressor
of the multifunctional printer are desired to have a throughput of
not less than a constant level, as a printer function, to satisfy
printing performance and to receive print information input from an
external device at a constant level of performance.
[0029] Although various lossless compression means are available,
when the throughput in compression process and decompression
process is constant, the amount of process resources required for
lossless compression is generally not proportional to the
compression rate but increases abruptly as the compression rate
rises while a value determined by the amount of data to be
compressed is used as a theoretical upper limit of the compression
rate. The process resources include the amount of calculation,
process steps, the amount of storage, etc. and are equivalent to
the circuit size of hardware. In other words, if lossless
compression process is simplified, the process resources required
for a constant throughput can be reduced, but the obtained
compression rate becomes relatively low. On the other hand, if
complicated lossless compression process is used, the amount of the
process resources required for a constant throughput becomes
relatively large, but the obtained compression rate can be made
closer to the theoretical upper limit.
[0030] In the above-mentioned JPEG, Lossless JPEG, JPEG-LS, JPEG
2000, etc., after data compression methods using two-dimensional
image correlations peculiar to the respective types of compression
process are performed, entropy coding is performed and two-step
data compression is performed. With respect to Huffman coding and
arithmetic coding serving as typical entropy coding, it is known
that the arithmetic coding can generally attain a higher
compression rate than the Huffman coding; on the other hand, if
compression rates approximately equal to each other are assumed and
the arithmetic coding and the Huffman coding are compared at the
same throughput, it is known that the arithmetic coding requires
more amount of process resources than the Huffman coding.
[0031] On the other hand, an image area to be subjected to lossless
compression contains text and vector images. Since the numbers of
texts, lines, etc. and the number of usable colors in these images
are not limited, it is possible to attain all data representations
for multilevel input images. For this reason, lossless compression
process is required to be able to compress multilevel input images
themselves. However, with respect to the Huffman coding and the
arithmetic coding, in the case of an adaptive encoding method in
which a generated code is changed particularly depending on the
input to be encoded, when the compression rate and the throughput
are constant, it is known that depending on the value range
(dynamic range) to be compressed, the required amount of the
process resources increases abruptly according to the power law or
at an exponential rate, instead of a linear functional
increase.
[0032] For the above-mentioned reasons, it is difficult to
simultaneously attain a throughput higher than a constant level, a
high image compression rate and a reduced amount of resources
required for process while the wide value range of an input image
is secured. Excellent lossless compression methods capable of
attaining these simultaneously have not yet been found at present
in any documents including Japanese Patent Application Laid-open
No. H03-104380 (1991), Japanese Patent Application Laid-open No.
2000-184205 and Japanese Patent Application Laid-open No.
2003-158739.
[0033] The present invention has been made with the aim of solving
the above problems, and it is an object of the invention to provide
an image compressing apparatus and an image compressing method
capable of efficiently performing compression process at a high
compression rate while degradation in image quality is reduced by
using lossy compression process and a plurality of different types
of lossless compression process in combination, to provide an image
decompressing apparatus and an image decompressing method capable
of efficiently performing decompression process for the image
compressed using the lossy compression process and the plurality of
different types of lossless compression process, and to an image
forming apparatus and a recording medium.
[0034] An image compressing apparatus according to the present
invention is an image compressing apparatus for compressing an
input image having a plurality of color components, comprising: a
first image segmentation section for segmenting said input image
into a image for lossless compression to be subjected to lossless
compression and a image for lossy compression to be subjected to
lossy compression, based on pixel identification information
indicating that respective pixels constituting said input image
belong to which of a plurality of areas including a text area and a
photograph area; a color determination section for determining
color information that is used when said image for lossless
compression is further segmented; a second image segmentation
section for segmenting said image for lossless compression
segmented by said first image segmentation section into a first
image for lossless compression containing said color information
determined by said color determination section and one or a
plurality of second images for lossless compression excluding said
first image for lossless compression; and an image compression
section for performing respectively different compression process
for each of said image for lossy compression segmented by said
first image segmentation section, said first image for lossless
compression and said second image for lossless compression
segmented by said second image segmentation section.
[0035] An image compressing apparatus according to the present
invention is characterized by further comprising: a frequency
distribution generating section for generating frequency
distribution of color information of respective pixels constituting
said image for lossless compression based on said image for
lossless compression, wherein said color determination section
determines the color information indicating a predetermined number
of colors whose frequency of occurrence is higher, based on the
frequency distribution generated by said frequency distribution
generating section.
[0036] An image compressing apparatus according to the present
invention is characterized in that said color determination section
determines the color information for respective predetermined areas
in said input image.
[0037] An image compressing apparatus according to the present
invention is characterized in that said first image segmentation
section segments said input image into said image for lossless
compression and said image for lossy compression according to a
predetermined condition.
[0038] An image compressing apparatus according to the present
invention is characterized in that said second image segmentation
section generates said first image for lossless compression by
replacing a pixel value of a pixel included in a pixel constituting
said image for lossless compression segmented by said first image
segmentation section and segmented into said second image for
lossless compression with a pixel value indicating a transparent
color.
[0039] An image compressing method according to the present
invention is an image compressing method for compressing an input
image having a plurality of color components, comprising: a first
image segmentation step for segmenting said input image into a
image for lossless compression to be subjected to lossless
compression and a image for lossy compression to be subjected to
lossy compression, based on pixel identification information
indicating that respective pixels constituting said input image
belong to which of a plurality of areas including a text area and a
photograph area; a color determination step for determining color
information that is used when said image for lossless compression
is further segmented; a second image segmentation step for
segmenting said image for lossless compression segmented at said
first image segmentation step into a first image for lossless
compression containing said color information determined at said
color determination step and one or a plurality of second images
for lossless compression excluding said first image for lossless
compression; and an image compression step for performing
respectively different compression process for each of said image
for lossy compression segmented by said first image segmentation
step, said first image for lossless compression and said second
image for lossless compression segmented by said second image
segmentation step.
[0040] An image decompressing apparatus according to the present
invention is an image decompressing apparatus for decompressing a
compressed image, comprising: an image decompression section for
performing respectively different decompression process for a lossy
compression image subjected to lossy compression after segmented
from one image, and a plurality of lossless compression images each
subjected to different types of lossless compression after
segmented from said one image; and an image composing section for
composing a plurality of images obtained by performing the
different types of decompression process by said image
decompression section, by superimposing the plurality of images in
a predetermined sequence.
[0041] An image decompressing method according to the present
invention is an image decompressing method for decompressing a
compressed image, comprising: an image decompression step for
performing respectively different decompression process for a lossy
compression image subjected to lossy compression after segmented
from one image, and a plurality of lossless compression images each
subjected to different types of lossless compression after
segmented from said one image; and an image composing step for
composing a plurality of images obtained by performing the
different types of decompression process at the image decompression
step, by superimposing the plurality of images in a predetermined
sequence.
[0042] An image forming apparatus according to the present
invention is an image forming apparatus comprising: the
above-mentioned image decompressing apparatus; and an image forming
section for forming an output image based on an image processed by
said image decompressing apparatus.
[0043] An image forming apparatus according to the present
invention is an image forming apparatus comprising: any one of the
above-mentioned image compressing apparatus; the above-mentioned
image decompressing apparatus; and an image forming section for
forming an output image based on the image processed by said image
decompressing apparatus, wherein said image decompressing apparatus
performs decompression process for an image compressed by said
image compressing apparatus.
[0044] A recording medium according to the present invention is a
recording medium on which a computer program for causing a computer
to perform compression process for an input image having a
plurality of color components are recorded so as to be readable by
said computer, said computer programs comprising:
a first image segmentation step for causing said computer to
segment said input image into a image for lossless compression to
be subjected to lossless compression and a image for lossy
compression to be subjected to lossy compression, based on pixel
identification information indicating that respective pixels
constituting said input image belong to which of a plurality of
areas including a text area and a photograph area; a color
determination step for causing said computer to determine color
information that is used when said image for lossless compression
is further segmented; a second image segmentation step for causing
said computer to segment said image for lossless compression
segmented at said first image segmentation step into a first image
for lossless compression containing said color information
determined at said color determination step and one or a plurality
of second images for lossless compression excluding said first
image for lossless compression; and an image compression step for
causing said computer to perform respectively different compression
process for each of said image for lossy compression segmented by
said first image segmentation step, said first image for lossless
compression and said second image for lossless compression
segmented by said second image segmentation step.
[0045] A recording medium according to the present invention is a
recording medium on which computer programs for causing a computer
to perform decompression process for a compressed image are
recorded so as to be readable by said computer, said computer
programs comprising: an image decompression step for causing said
computer to perform respectively different decompression process
for a lossy compression image subjected to lossy compression after
segmented from one image, and a plurality of lossless compression
images each subjected to different types of lossless compression
after segmented from said one image; and an image composing step
for causing said computer to compose a plurality of images obtained
by the different types of decompression process at said image
decompression step by superimposing the plurality of images in a
predetermined sequence.
[0046] In the present invention, an input image is segmented into a
image for lossless compression and image for a lossy compression
based on the pixel identification information of respective pixels
constituting the input image. In addition, the image for lossless
compression is segmented into a first image for lossless
compression containing predetermined color information and one or a
plurality of second images for lossless compression excluding the
first image for lossless compression. Different types of
compression process are performed for the segment image for lossy
compression, first image for lossless compression and second image
for lossless compression. In other words, the input image is
segmented into areas respectively suited for lossless compression
and lossy compression, and lossless compression or lossy
compression is performed for each area, whereby efficient
compression process can be performed. Furthermore, the image for
lossless compression is segmented into the first image for lossless
compression containing the predetermined color information and the
second image for lossless compression other than the first image
for lossless compression, and different types of lossless
compression are performed for the first and second images for
lossless compression, whereby more efficient compression process
can be performed.
[0047] In the present invention, the color information that is used
when the image for lossless compression is segmented into the first
image for lossless compression and the second image for lossless
compression other than the first image for lossless compression is
determined based on the frequency distribution of the color
information of respective pixels contained in the image for
lossless compression. The ratio of the first image for lossless
compression in the lossless compression image can be raised by
segmenting the image containing the color information appearing
frequently into the first image for lossless compression. Hence,
for example, when lossless compression process expected to provide
a high compression rate is performed for the first image for
lossless compression, a high compression rate can be attained. In
addition, the first image for lossless compression can be prevented
from increasing and efficient lossless compression can be attained
by limiting the amount of the color information segmented into the
first image for lossless compression to a predetermined number.
More appropriate compression process can be performed by performing
compression process for the first image for lossless compression
and the second image for lossless compression in consideration of
image compression rate and resources required for the compression
process.
[0048] In the present invention, the color information according to
which the image for lossless compression is segmented into the
first image for lossless compression and the second image for
lossless compression is made different for each predetermined area
in the input image. Hence, even an input image having local areas
in which different color information is used, a wider range in the
image for lossless compression can be segmented into the first
image for lossless compression. As a result, for example, when
lossy compression process expected to provide a high compression
rate is performed for the first image for lossless compression, a
high compression rate can be attained, and efficient compression
process can be attained.
[0049] In the present invention, the input image is segmented into
a image for lossless compression and a image for lossy compression
according to a preset condition. Hence, for example, when a
predetermined condition is set and the entire input image is
segmented into the image for lossless compression, the input image
is reproduced only from the image for lossless compression, whereby
the degradation in image quality before and after compression
process can be prevented.
[0050] In the present invention, when the first image for lossless
compression containing the predetermined color information is
segmented from the image for lossless compression segmented from
the input image, the pixel values of the pixels included in the
pixels constituting the image for lossless compression and
segmented into the second image for lossless compression are
replaced with a pixel value indicating a transparent color. Hence,
the first image for lossless compression and the second image for
lossless compression are composed only by superimposing the first
image for lossless compression on the second image for lossless
compression, whereby the image for lossless compression can be
restored easily.
[0051] In the present invention, after different types of
decompression process are performed for the lossy compression image
segmented from one image and subjected to lossy compression and for
a plurality of lossless compression images compressed by different
types of lossless compression process, the images are superimposed
in a predetermined sequence so as to be composed. Hence, a
decompressed image can be generated easily by decompressing and
superimposing the compressed images, without requiring mask images
or the like for the decompressed images.
[0052] According to the present invention, compression process can
be performed efficiently at a high compression rate by combing
lossless compression process with lossy compression process and
also by combing a plurality of different types of lossless
compression process with each other. In particular, compression
process can be performed efficiently for an image having numerous
objects, such as a text (a character), a graphic (diagram) and a
photograph, while the reproducibility of edges and the reusability
of the image are maintained. Furthermore, decompression process can
be performed efficiently for an image compressed using lossy
compression process and a plurality of different types of lossless
compression process.
[0053] The above and further objects and features will more fully
be apparent from the following detailed description with
accompanying drawings.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0054] FIG. 1 is a block diagram showing the internal configuration
of a conventional printer;
[0055] FIG. 2 is a block diagram showing the internal configuration
of another conventional printer;
[0056] FIG. 3 is a block diagram showing the internal configuration
of still another conventional printer;
[0057] FIG. 4 is a block diagram showing the configuration of an
image forming apparatus according to Embodiment 1;
[0058] FIG. 5 is a block diagram showing the configuration of an
image compression section;
[0059] FIG. 6 is a characteristic graph showing the relationship
between the number of color codes used in an image and an image
covering ratio;
[0060] FIGS. 7A and 7B are schematic views for explaining process
of segmenting continuous tone bit map image data;
[0061] FIGS. 8A, 8B and 8C are schematic views for explaining
process of segmenting the continuous tone bit map image data;
[0062] FIG. 9 is a flowchart showing the procedure of the image
compression process performed by the image compression section;
[0063] FIG. 10 is another flowchart showing the procedure of the
image compression process performed by the image compression
section;
[0064] FIG. 11 is a block diagram showing the configuration of an
image decompression section;
[0065] FIG. 12 is a flowchart showing the procedure of the image
decompression process performed by the image decompression
section;
[0066] FIG. 13 is another flowchart showing the procedure of the
image decompression process performed by the image decompression
section;
[0067] FIG. 14 is a table showing the effects obtained when the
present invention is applied;
[0068] FIG. 15 is a block diagram showing the configuration of the
main section of a PC according to Embodiment 3; and
[0069] FIG. 16 is a block diagram showing the configuration of an
image forming apparatus according to Embodiment 3.
DETAILED DESCRIPTION
[0070] An image compressing apparatus, an image compressing method,
an image decompressing apparatus, an image decompressing method, an
image forming apparatus, computer programs and a recording medium
according to the present invention will be described below in
detail based on the drawings showing the embodiments thereof.
Embodiment 1
[0071] An image forming apparatus according to Embodiment 1 will be
described below. FIG. 4 is a block diagram showing the
configuration of an image forming apparatus 100 according to
Embodiment 1. The image forming apparatus 100 (for example, a
digital color multi-function peripheral) according to Embodiment 1
is equipped with an image input section 11, an image process
section 10, an image output section 12, an external interface
section 13, an operation panel 14, etc. The operations of these
components provided for the image forming apparatus 100 are
controlled by a CPU (central process unit), not shown.
[0072] The image forming apparatus 100 performs predetermined image
process for image data obtained from the image input section 11 or
the external interface section 13 using the image process section
10. Furthermore, the image forming apparatus 100 outputs (prints)
images based on the processed image data using the image output
section 12 or outputs the processed image data to an external
device from the external interface section 13.
[0073] The image input section 11 is a scanner equipped with a
light source for irradiating reading light to a document, a CCD
(charged coupled device) line sensor, etc. The image input section
11 converts the light reflected by the document into an analog R, G
and B (R: red, G: green, B: blue) signal (RGB reflectivity signal).
Since the CCD line sensor is used as an image sensing device, the
image input section 11 reads a two-dimensional image by performing
reading while the light source is used for scanning in a direction
(sub scanning direction) orthogonal to the longitudinal direction
(main scanning direction) of the CCD line sensor.
[0074] In addition, the image input section 11 has an A/D
(analog/digital) converter. The image input section 11 performs A/D
conversion process for the obtained analog image signal and
converts the signal into an 8-bit digital image signal, for
example. Hence, the image input section 11 generates scanner image
data (raster data) having R, G and B color components and outputs
the data to the image process section 10.
[0075] The image output section 12 is an electrophotographic type
or ink-jet type printer (an image forming section). Based on the
print image data received from the image process section 10, the
image output section 12 forms an image on a sheet of recording
paper or OHP film or the like, thereby outputting the image.
[0076] The operation panel 14 is equipped with an operation section
for accepting operations, such as the setting of the operation mode
of the image forming apparatus 100, a display section, such as a
liquid crystal display, etc.
[0077] The external interface section 13 having a wired or wireless
external connection function receives the PDL data from an external
PC or the like via a wired or wireless communication path and
outputs the data to the image process section 10. Furthermore, the
external interface section 13 receives the transmission image data
generated by the image process section 10 and transmits the data to
the external PC or the like. In Embodiment 1, a configuration is
described in which the external interface section 13 receives the
PDL data written in PDL, such as PostScript, from the external PC
or the like. However, the data is not limited to data according to
the PDL format, but may be data specific to the image forming
apparatus 100, for example, provided that the data is used to
instruct the image forming apparatus 100 on printing process or
transmission process.
[0078] The image process section 10 is equipped with a scanner
image process section 1, an RIP process section 2, an input
selector 3, an image compression section 4 (an image compressing
apparatus), a storage section 5, an image decompression section 6
(an image decompressing apparatus), an output selector 7, a print
image process section 8, a transmission image process section 9,
etc.
[0079] The scanner image process section 1 performs shading
correction process to eliminate various kinds of distortions caused
in scanner image data input from the image input section 11 due to
the configurations of the lighting system, the image focusing
system, the image sensing system, etc. of the image input section
11. Furthermore, the scanner image process section 1 performs
visual sensitivity correction using LUTs (look-up tables) prepared
for the respective RGB signals to correct the difference between
the sensitivity characteristics of the CCD line sensor serving as
an image sensing device and the visual sensitivity characteristics
of the human. The scanner image process section 1 supplies
corrected scanner image data to the input selector 3. The corrected
scanner image data has an image size determined depending on the
output resolution and the output size of the image output section
12. Respective pixels of the scanner image data are represented by
256 gray scales in which each color component of RGB can be
represented by 8 bits.
[0080] The RIP process section 2 interprets the PDL data input from
the external PC or the like via the external interface section 13
according to the PDL language specifications and generates
continuous tone bit map image data (raster data) having RBG color
components. The continuous tone bit map image data has an image
size determined depending on the output resolution and the output
size of the image output section 12. Respective pixels of the
continuous tone bit map image data are represented by 256 gray
scales in which each color component of RGB can be represented by 8
bits.
[0081] Furthermore, the RIP process section 2 determines whether
each pixel of the continuous tone bit map image data is contained
in any one of a text area (character area), a vector area (graphic
area), a photograph area and the other area (unclassified area).
The RIP process section 2 generates pixel identification
information data having pixel identification information
representing each area determined for each pixel. The RIP process
section 2 supplies the generated continuous tone bit map image data
and the generated pixel identification information data to the
input selector 3. The other area is an area in which nothing is
drawn, for example.
[0082] The RIP process section 2 may process a specific format
peculiar to the apparatus, in addition to the general-purpose
language, such as PDL, or process these combinations. Examples of
the specific format peculiar to the apparatus include a format in
which a bit map image representing a whole page or a part of a page
is compressed and delivered or delivered without compression, and a
format in which a bit map image representing a whole page or a part
of a page is divided into two parts, that is, text information and
the other information, and each of the parts is compressed and
delivered or delivered without compression.
[0083] Furthermore, another format is available in which a bit map
image representing a whole page is divided more finely into three
parts, that is, text information, graphic information and
photograph information and these are compressed and delivered or
delivered without compression. Moreover, still another format is
available in which a bit map image representing a part of a page is
compressed and delivered or delivered without compression and
supplementary information representing that each portion thereof is
a text, a graphic or a photograph is also delivered at the same
time, and yet still another format is available in which text
information, graphic information and photograph information are not
designated as bit map images but as commands. Although the
above-mentioned formats are taken as examples of the specific
format peculiar to the apparatus, the format is not limited to
these.
[0084] Moreover, the RIP process section 2 is not limited to the
generation of the continuous tone bit map image data having three
color components of RGB, but may generate continuous tone bit map
image data represented by the luminance tone of only one color.
Still further, when PDL data or data having a specific format
peculiar to the apparatus, supporting multicolor representation,
for example, by CMYK (C: cyan, M: magenta, Y: yellow, K: black) of
PostScript or the like, is used, the RIP process section 2 should
only generate continuous tone bit map image data having four color
components, such as CMYK. However, the color space representing the
continuous tone bit map image data is not limited to the
above-mentioned RGB or CMYK.
[0085] When pixel identification information data is generated from
PDL data, the RIP process section 2 first sets values indicating an
unclassified area as the initial values of the pixel identification
information corresponding to all pixels before generating the
continuous tone bit map image data. When drawing respective objects
(drawn elements that can be interpreted from instructions supported
by PDL, such as texts, graphics and images) in the continuous tone
bit map image data while interpreting PDL and when the objects have
been drawn as text information by using PDL instructions as
information for making decisions, the RIP process section 2 changes
the pixel identification information corresponding to the drawn
pixels to values indicating a text area. In addition, when the
objects have been drawn as graphic information, the RIP process
section 2 changes the pixel identification information
corresponding to the drawn pixels to values indicating a vector
area.
[0086] Furthermore, when the PDL data contains an area compressed
using a lossy compression method, such as JPEG or JPEG 2000, the
RIP process section 2 changes the pixel identification information
corresponding to the pixels drawn based on this area to values
indicating a photograph area. The RIP process section 2 performs
the above-mentioned process for all the pixels in the continuous
tone bit map image data generated from the PDL data, thereby
generating pixel identification information data having the pixel
identification information corresponding to respective pixels.
[0087] In Embodiment 1, the pixel identification information data
contains four kinds of pixel identification information, that is, a
text area, a vector area, a photograph area and the other area
(unclassified area). However, the classification method is not
limited to the above-mentioned method of performing classification
depending on the kind of object. For example, it may be possible to
use various classification methods depending on the designation by
the user or the operation mode of the image forming apparatus 100.
More specifically, when large size texts have been drawn based on
text information, values indicating a photograph area instead of a
text area may be set in the pixel identification information
corresponding to the drawn pixels or values indicating a large text
area different from an ordinary text area may also be set.
[0088] In addition, also in a vector area, when a graphic having
height and width values larger than constant values have been drawn
based on graphic information, values indicating a photograph area
instead of a vector area may be set in the pixel identification
information corresponding to the drawn pixels or values indicating
a large vector area different from an ordinary vector area may also
be set. In this way, classification depending on the size of the
object can be performed by setting pixel identification information
different depending on whether the size of the same kind of object
is a predetermined size or larger.
[0089] Since the pixel identification information data is used as
reference information when the print image process section 8 in the
subsequent stage performs color conversion and halftone generation
for image data, it may be possible to use pixel identification
information in which text and vector areas are divided more finely.
For example, a text area may be finely divided into a black text
area and a color text area to distinguish whether a text area is an
achromatic text area or a chromatic text area. Furthermore, in a
similar way, a vector area may be finely divided into a black
vector area and a color vector area to distinguish whether a
graphic area is an achromatic graphic area or a chromatic graphic
area. Hence, when the print image process section 8 performs color
conversion and halftone generation, monochromatization (black)
process or the like can be performed easily.
[0090] In addition, the pixel identification information data
should only contain pixel identification information corresponding
to the respective pixels of the continuous tone bit map image data
and is not necessarily required to contain the same number of the
pixel identification information as the number of pixels of the
continuous tone bit map image data. For example, in the continuous
tone bit map image data, totally 64 pixels comprising 8 pixels in
the main scanning direction and 8 pixels in the sub scanning
direction are regarded as one block, and one piece of pixel
identification information may be set in one block. In this case,
the same pixel identification information should only be assigned
to all the pixels contained in each block when the pixel
identification information data is referred to. The amount of the
pixel identification information data can be reduced by setting one
piece of pixel identification information in one block as described
above.
[0091] The input selector 3 supplies either the scanner image data
from the scanner image process section 1 or the continuous tone bit
map image data from the RIP process section 2 to the image
compression section 4 in the subsequent stage. When the RIP process
section 2 has supplied the pixel identification information data
together with the continuous tone bit map image data to the input
selector 3 and the input selector 3 supplies the continuous tone
bit map image data to the image compression section 4, the input
selector 3 also supplies the pixel identification information data
to the image compression section 4. The input selector 3 has a
temporary buffer memory capable of storing the scanner image data,
the continuous tone bit map image data and the pixel identification
information data. After receiving one page of input, the input
selector 3 performs store-and-forward operation to supply the data
to the image compression section 4.
[0092] Usually, the input selector 3 preferentially selects the
input from the scanner image process section 1 and supplies the
input to the image compression section 4. For example, when the
input from the scanner image process section 1 is started while the
input selector 3 receives data from the RIP process section 2, the
input selector 3 stores the data being received at present from the
RIP process section 2 in the temporary buffer memory and then
interrupts subsequent data reception from the RIP process section
2. Then, the input selector 3 starts data reception from the
scanner image process section 1 and stops data reception from the
RIP process section 2 until data reception from the scanner image
process section 1 is completed. While the input selector 3 stops
data reception from the RIP process section 2, the RIP process
section 2 temporarily stops the process of generating the
continuous tone bit map image data and the pixel identification
information data.
[0093] The image compression section 4 receives either the scanner
image data or the continuous tone bit map image data and the pixel
identification information data supplied from the input selector 3
and then performs data compression. The configuration of the image
compression section 4 and the process performed by the image
compression section 4 will be described later in detail.
[0094] The storage section 5 is constituted by a hard disk drive
(HDD) or a nonvolatile storage unit, such as a flash memory, and
receives data compressed by the image compression section 4 and
stores the data. Furthermore, the compressed data stored in the
storage section 5 is read by the image decompression section 6 in
the subsequent stage.
[0095] The image decompression section 6 decompresses the
compressed data read from the storage section 5 and supplies the
decompressed scanner image data or the decompressed continuous tone
bit map image data and pixel identification information data to the
output selector 7. The configuration of the image decompression
section 6 and the process performed by the image decompression
section 6 will be described later in detail.
[0096] The output selector 7 supplies the scanner image data or the
continuous tone bit map image data and the pixel identification
information data, decompressed by the image decompression section
6, to the print image process section 8 or the transmission image
process section 9 according to instructions from the CPU. More
specifically, when the image data is to be printed out by the image
output section 12, the image data is supplied to the print image
process section 8. When the image data is transmitted to the
external PC via the external interface section 13, the image data
is supplied to the transmission image process section 9.
[0097] The print image process section 8 receives the data (the
scanner image data, or the continuous tone bit map image data and
the pixel identification information data) supplied from the output
selector 7 and performs image process suited for the image output
section 12. In addition, the print image process section 8
generates halftone image data for printing (print image data) and
supplies the image data to the image output section 12. The print
image process section 8 is equipped with a color conversion process
section 8a and a halftone generation section 8b. The data supplied
from the output selector 7 to the print image process section 8 is
input to the color conversion process section 8a.
[0098] The color conversion process section 8a converts the data
(the scanner image data or the continuous tone bit map image data)
having three color components of RGB into continuous tone bit map
image data having four colors of CMYK for printing. The halftone
generation section 8b converts the continuous tone bit map image
data for printing into halftone image data having four colors of
CMYK for printing. On receiving the pixel identification
information data from the output selector 7, the halftone
generation section 8b may generate halftone bit map image data for
printing using a halftone generating method different for each
piece of the pixel identification information of the pixel
identification information data.
[0099] For example, when a dither method is used as a halftone
generating method, a threshold matrix used for the pixels of a text
area and a vector area in the dither method is configured so that
resolution is regarded more important than tone reproducibility,
and a threshold matrix used for the pixels of a photograph area in
the dither method is configured so that tone reproducibility is
regarded more important than resolution. Furthermore, when an error
diffusion method is used as a halftone generating method, error
diffusion is performed only for neighboring pixels with respect to
the text area and the vector area, and error diffusion is performed
for wider neighboring pixels with respect to the photograph
area.
[0100] The transmission image process section 9 receives the data
(the scanner image data or the continuous tone bit map image data
and the pixel identification information data) supplied from the
output selector 7 and performs image process suited for a
transmission image. In addition, the transmission image process
section 9 generates JPEG compressed image data (transmission image
data) and supplies the image data to the external interface section
13. The image process suited for the transmission image is, for
example, filter process in which a filter different for each piece
of pixel identification information of the pixel identification
information data is used. The filter process serves as edge
enhancement process (sharpening process) for the pixels of a text
area.
[0101] The transmission image process section 9 is equipped with a
zoom process section 9a, a JPEG compression section 9b. The data
supplied from the output selector 7 to the transmission image
process section 9 is input to the zoom process section 9a. The data
(the scanner image data or the continuous tone bit map image data)
having three color components of RGB has an image size and an
output resolution suited for the image output section 12. Hence,
the zoom process section 9a performs zooming so that the input data
has an image size and a resolution suited for transmission to the
external PC via the external interface section 13, to generate the
continuous tone bit map image data. The JPEG compression section 9b
compresses the zoomed continuous tone bit map image data to obtain
image data of a format (JPEG format) according to JPEG File
Interchange Format (JFIF) and supplies the image data to the
external interface section 13.
[0102] Next, the image compression section 4 will be described
below. The scanner image data output from the scanner image process
section 1 or the continuous tone bit map image data and the pixel
identification information data output from the RIP process section
2 are input to the image compression section 4. In order that the
image data output from the scanner image process section 1 is
distinguished from the image data output from the RIP process
section 2, the data output from the scanner image process section 1
has been referred to as the scanner image data. However, the
scanner image data is also continuous tone bit map image data, and
the continuous tone bit map image data containing the scanner image
data will also be referred to as continuous tone bit map image data
in the following descriptions.
[0103] The image compression section 4 generates, based on the
input data, one kind of lossy compression image data, two kinds of
lossless compression image data (first lossless compression image
data and second lossless compression image data), index color
information and compressed pixel identification information data at
a maximum. The image compression section 4 outputs the five kinds
of the generated data at a maximum as one set and stores them in
the storage section 5. When the data output from the scanner image
process section 1 through the input selector 3 has been input to
the image compression section 4, the pixel identification
information data is not input to the image compression section 4.
Furthermore, since the image compression section 4 performs only
lossy compression process for the data output from the scanner
image process section 1, only lossy compression image data is
generated in this case.
[0104] FIG. 5 is a block diagram showing the configuration of the
image compression section 4. The image compression section 4 is
equipped with a buffer memory 41, a pixel statistics section 42
(color determination section, frequency distribution generating
section), a photograph image segmentation section 43 (first image
segmentation section), an index image segmentation section 44
(first image segmentation section, second image segmentation
section), a lossy compression section 45 (image compression
section), a first lossless compression section 46 (image
compression section), a second lossless compression section 47
(image compression section), an identification information
compression section 48, etc.
[0105] The image compression section 4 according to Embodiment 1
compresses continuous tone bit map image data to be input, by using
the line data of 8 lines in the main scanning direction as one
process unit. This is because the lossy compression section 45
according to Embodiment 1 is configured so as to perform lossy
compression process according to the JPEG method. Hence, when the
lossy compression section 45 is configured so as to perform
compression process other than that according to the JPEG method,
one process unit should only be determined in consideration of the
characteristics of the compression methods used for the lossy
compression section 45, the first lossless compression section 46
and the second lossless compression section 47, the parameters, the
capacity of the buffer memory 41, effects on image quality due to
compression process, compression efficiency, etc.
[0106] The continuous tone bit map image data and the pixel
identification information data output from the input selector 3
are input to the pixel statistics section 42. When the data output
from the scanner image process section 1 through the input selector
3 has been input to the image compression section 4, only the
continuous tone bit map image data is input to the pixel statistics
section 42. In this case, the pixel statistics section 42 performs
process assuming that the pixel identification information data, in
which the pixel identification information corresponding to all the
pixels in the continuous tone bit map image data to be input
contains values indicating an unclassified area, has been
input.
[0107] The pixel statistics section 42 generates block segmentation
information and index color information based on the continuous
tone bit map image data and the pixel identification information
data. By regarding totally 64 pixels comprising 8 pixels in the
main scanning direction and 8 pixels in the sub scanning direction
as one block in the continuous tone bit map image data, the block
segmentation information is used as information indicating whether
lossy compression or lossless compression is performed for each
block. More specifically, the block segmentation information is
one-bit information per block. The value 0 of the block
segmentation information indicates that the corresponding block is
a block in which lossy compression is performed, and the value 1
indicates that the corresponding block is a block in which lossless
compression is performed. Hence, in Embodiment 1, block
segmentation information for 1/8 times the number of the pixels
contained in one line in the main scanning direction is generated
when process (process for the line data of 8 lines) is performed
once.
[0108] As described above, in Embodiment 1, the lossy compression
section 45 is configured so as to perform lossy compression process
according to the JPEG method, and the minimum block size during the
DCT compression process according to the JPEG method is totally 64
pixels comprising 8 pixels in the main scanning direction and 8
pixels in the sub scanning direction. Hence, the pixel statistics
section 42 also generates block segmentation information by using
64 pixels as one process unit similarly, but the process unit is
not limited to this. For example, when JPEG 2000 is used instead of
JPEG as a lossy compression method, the pixel statistics section 42
may generate block segmentation information in each tile unit or
each pixel unit. Hence, as in the determination of one process unit
in the image compression section 4, the generation unit of the
block segmentation information should only be determined in
consideration of the characteristics of the compression methods
used for the lossy compression section 45, the first lossless
compression section 46 and the second lossless compression section
47, the parameters, the capacity of the buffer memory 41, effects
on image quality due to compression process, compression
efficiency, etc.
[0109] The image compression section 4 according to Embodiment 1
performs lossless compression for text and vector areas in an
image. Hence, the pixel statistics section 42 refers to the pixel
identification information respectively corresponding to 64 pixels
contained for each block, and when the pixel statistics section 42
judges that at least one pixel of the text and vector areas is
contained in the block, the pixel statistics section 42 considers
that lossless compression is performed in this block, and value 1
is set in the block segmentation information of this block. On the
other hand, when no pixel of the text and vector areas is contained
in the block, the pixel statistics section 42 considers that lossy
compression is performed in this block, and value 0 is set in the
block segmentation information of this block.
[0110] When the continuous tone bit map image data from the scanner
image process section 1 has been input to the pixel statistics
section 42, the pixel statistics section 42 considers that the
pixel identification information data, in which the pixel
identification information corresponding to all the pixels contains
values indicating an unclassified area, has been input. In this
case, the pixel statistics section 42 sets values 0 in the block
segmentation information of all the blocks. The method for setting
the block segmentation information may be changed depending on the
designation by the user or the operation mode of the image forming
apparatus 100. For example, when degradation of image quality due
to image compression is desired to be eliminated, a function for
setting values 1 in the block segmentation information of all the
blocks may be provided, instead of using the pixel identification
information data.
[0111] The index color information is information that is used to
segment an area to be subjected to lossless compression by the
first lossless compression section 46 from an area to be subjected
to lossless compression in the continuous tone bit map image data,
in one process unit (the line data of 8 lines). In the following
descriptions, in the continuous tone bit map image data, an area to
be subjected to lossy compression is referred to as bit map image
data for lossy compression. In the area to be subjected to lossless
compression, the area to be subjected to lossless compression by
the first lossless compression section 46 is referred to as index
image data for lossless compression. Furthermore, in the area to be
subjected to lossless compression, an area other than the index
image data for lossless compression, more specifically, the area to
be subjected to lossless compression by the second lossless
compression section 47, is referred to as bit map image data for
lossless compression.
[0112] In Embodiment 1, when the index image data for lossless
compression is segmented from an area to be subjected to lossless
compression in the continuous tone bit map image data, the color
information (color code) of each pixel is used. More specifically,
in the color information of each pixel included in areas to be
subjected to lossless compression in the continuous tone bit map
image data, that is, in the text and vector areas, the pixels of
color information used frequently are segmented as the index image
data for lossless compression. Hence, the index color information
indicates the color information of the pixels to be segmented as
the index image data for lossless compression. In Embodiment 1,
index color information indicating the color information of 15
colors is generated in descending order of frequency of occurrence,
for one process unit (the line data of 8 lines).
[0113] The pixel statistics section 42 first creates a color
histogram for an area to be subjected to lossless compression in
the continuous tone bit map image data. The pixel statistics
section 42 refers to the pixel identification information data,
counts the number of pixels classified by each color value with
respect to the pixels in the text and vector areas in one process
unit (the line data of 8 lines), and creates a color histogram
(frequency distribution). The pixel statistics section 42 performs
descending sorting for the color histogram created in one process
unit in descending order of the number of pixels. Then, the pixel
statistics section 42 determines 15 colors (index colors) in
descending order of the number of pixels and assigns the color
codes of 15 values, 1 to 15, to the determined 15 colors,
respectively. The pixel statistics section 42 correlates the color
values of the 15 colors with the color codes assigned to the
respective color values, thereby generating the index color
information. Color code 0 is used to indicate a transparent color
as described later.
[0114] The continuous tone bit map image data generated from
general PDL data by the image forming apparatus 100 will be
described herein. FIG. 6 is a characteristic graph showing the
relationship between the number of color codes used in an image and
an image covering ratio. The horizontal axis in FIG. 6 represents
the number of color codes used in an image, showing the range of
numerals 1 to 1024. The vertical axis in FIG. 6 represents the
pixel covering ratio with respect to the number of color codes for
all the pixels in the text and vector areas contained in each
process unit, by regarding the line data of 8 lines as one process
unit in the continuous tone bit map image data. For example, when
the number of color codes is 1, the covering ratio is appropriately
50%. This indicates that pixels of appropriately 50% among the
pixels contained in the line data of 8 lines and corresponding to
the above-mentioned condition have the same color (one color).
[0115] In other words, as shown in FIG. 6, the number of colors
used in the text and vector areas in the continuous tone bit map
image data generated from general PDL data is usually not large.
Even when the number of color codes is 15, the covering ratio is
90% or more. Hence, in Embodiment 1, the wide range of the area to
be subjected to lossless compression in the continuous tone bit map
image data can be segmented as index image data for lossless
compression by using the color codes of 15 colors used frequently
in one process unit as index color information.
[0116] The first lossless compression section 46 is configured so
as to perform compression process that is expected to obtain a
compression rate higher than that obtained by the compression
process of the second lossless compression section 47 as described
later in detail. Hence, the compression efficiency of the entire
image can be improved by segmenting the wide range of the area to
be subjected to lossless compression in the continuous tone bit map
image data as index image data for lossless compression.
[0117] The method for determining the index colors based on a color
histogram by the pixel statistics section 42 is not limited to the
method based on the number of pixels (frequency of occurrence)
described above. For example, the colors used in the text area may
be preferentially selected as index colors. In this case, for
example, flags indicating colors to be used in the text area are
prepared for respective color values in the color histogram, and
the flags are set for the color values used in the text area when
the pixel statistics section 42 creates a color histogram. Then,
when the pixel statistics section 42 performs sorting for the
created color histogram, the color values for which the flags have
been set are sorted preferentially so as to be ranked high. As a
result, the color values for which the flags have been set, that
is, the colors used in the text area are preferentially selected as
index colors.
[0118] In the text and vector areas in the continuous tone bit map
image data, it is expected that the number of the pixels included
in the ground color of an image is large. Furthermore, since the
pixels included in the ground color of the image are not required
to be compressed at a high compression rate, it may be possible
that, for example, white that is used as the ground color of the
image at a high possibility, is not selected as an index color. In
addition, since it is also expected that black is used frequently
for texts, it may be possible that black is selected preferentially
as an index color.
[0119] Furthermore, although 15 colors have been determined as
index colors in Embodiment 1, when the number of colors used is
less than 16 in total as a result of the sorting performed for a
color histogram, all the color codes ranging from 0 to 15 may be
assigned to the respective colors used. In other words, in this
case, the respective color codes are used, assuming that color code
0 does not indicate a transparent color. When only 16 colors are
used in an area to be subjected to lossless compression in
continuous tone bit map image data, the entire area to be subjected
to lossless compression is compressed by the first lossless
compression section 46, whereby the compression rate is further
improved.
[0120] The pixel statistics section 42 outputs the generated block
segmentation information to the photograph image segmentation
section 43 and the index image segmentation section 44 and outputs
the generated index color information to the index image
segmentation section 44.
[0121] The buffer memory 41 has a memory capacity capable of
storing the line data of 8 lines in continuous tone bit map image
data represented by 24 bits (8 bits.times.3 colors) in one pixel.
The buffer memory 41 sequentially stores the line data of 8 lines
in the continuous tone bit map image data input from the input
selector 3. The pixel statistics section 42 generates block
segmentation information and index color information for the line
data of 8 lines stored in the buffer memory 41. Hence, the buffer
memory 41 outputs the stored line data of 8 lines to the photograph
image segmentation section 43 and the index image segmentation
section 44 at the timing when the pixel statistics section 42
outputs the block segmentation information and the index color
information.
[0122] The continuous tone bit map image data (the line data of 8
lines) from the buffer memory 41 and the block segmentation
information from the pixel statistics section 42 are input to the
photograph image segmentation section 43. Based on the block
segmentation information, the photograph image segmentation section
43 generates bit map image data for lossy compression in which each
color of RGB is represented by 8 bits (256 gray scales) from the
continuous tone bit map image data and outputs the bit map image
data to the lossy compression section 45. More specifically, the
photograph image segmentation section 43 specifies a block in which
the value of the block segmentation information is 0 and extracts
the pixel values of the pixels contained in the specified block. In
addition, the photograph image segmentation section 43 masks the
pixel values of the pixels contained in a block, for example, in
which the block segmentation information is not 0, using a
predetermined value and generates bit map image data for lossy
compression.
[0123] The continuous tone bit map image data (the line data of 8
lines) from the buffer memory 41 and the block segmentation
information and the index color information from the pixel
statistics section 42 are input to the index image segmentation
section 44. Based on the block segmentation information and the
index color information, the index image segmentation section 44
generates index image data for lossless compression in which each
pixel is represented by 4 bits (16 values) and bit map image data
for lossless compression in which each color of RGB is represented
by 8 bits (256 gray scales) from the continuous tone bit map image
data.
[0124] More specifically, the index image segmentation section 44
specifies a block in which the value of the block segmentation
information is 1 and extracts the pixel values of the pixels
contained in the specified block. In addition, the index image
segmentation section 44 judges whether the color values of
respective pixels indicated by the extracted pixel values are equal
to the color values contained in the index color information. When
the color values of the respective pixels are equal to the color
values contained in the index color information, the index image
segmentation section 44 replaces the pixel values with color codes
(1 to 15) correlated with the color values and uses the color codes
as the pixel values of the index image data for lossless
compression.
[0125] When the color values of the respective pixels indicated by
the extracted pixel values are not equal to the color values
contained in the index color information, the index image
segmentation section 44 replaces the pixel values with color code 0
and uses color code 0 as the pixel value of the index image data
for lossless compression. In the index image data for lossless
compression, pixel value 0 is used to indicate a transparent color
in the following process. Hence, in the areas to be subjected to
lossless compression in the continuous tone bit map image data, the
areas in which index colors are used are extracted, and index image
data for lossless compression is generated.
[0126] Furthermore, the index image segmentation section 44
extracts the pixel values of pixels in which the color values
indicated by the pixel values are not equal to the color values
contained in the index color information among the pixels contained
in blocks in which the value of the block segmentation information
is 1 and uses the pixel values as the pixel values of the bit map
image data for lossless compression. Hence, in the areas to be
subjected to lossless compression in the continuous tone bit map
image data, the areas other than the areas in which the index
colors are used are extracted, and bit map image data for lossless
compression is generated.
[0127] The index image segmentation section 44 outputs the
generated index image data for lossless compression to the first
lossless compression section 46 and outputs the generated bit map
image data for lossless compression to the second lossless
compression section 47.
[0128] FIGS. 7A, 7B, 8A, 8B and 8C are schematic views for
explaining process of segmenting continuous tone bit map image
data. FIG. 7A shows an example of continuous tone bit map image
data, showing a part of the line data of 8 lines. The vertical
lines in FIG. 7A indicate the boundaries of blocks in the line data
of 8 lines, wherein one block is regarded as totally 64 pixels
comprising 8 pixels in the main scanning direction and 8 pixels in
the sub scanning direction. Furthermore, the respective hatched
areas in FIG. 7A represent colors different from one another.
[0129] FIG. 7B shows an example of pixel identification information
data corresponding to the continuous tone bit map image data shown
in FIG. 7A. FIG. 7B also shows an example in which each pixel is
contained in any one of a text area, a vector area, a photograph
area and an unclassified area. FIG. 8A shows an example of bit map
image data for lossy compression generated from the continuous tone
bit map image data shown in FIG. 7A based on the pixel
identification information data shown in FIG. 7B. Furthermore,
FIGS. 8B and 8C respectively shows examples of index image data for
lossless compression and bit map image data for lossless
compression generated from the continuous tone bit map image data
shown in FIG. 7A based on the pixel identification information data
shown in FIG. 7B.
[0130] As shown in FIG. 8A, only one block in which all the pixels
are contained in the photograph area is contained in the bit map
image data for lossy compression. In FIG. 8A, only the rightmost
block is extracted from the continuous tone bit map image data, and
the pixel values of the pixels of the other blocks are masked with
a predetermined value, for example. As shown in FIG. 8B, among the
pixels in blocks containing at least one pixel contained in the
text area or the vector area, the pixels having color values
corresponding to the index colors are contained in the index image
data for lossless compression.
[0131] As shown in FIG. 8C, image data obtained by eliminating the
bit map image data for lossy compression shown in FIG. 8A and the
index image data for lossless compression shown in FIG. 8B from the
continuous tone bit map image data shown in FIG. 7A is contained in
the bit map image data for lossless compression. In other words,
among the pixels in blocks containing at least one pixel contained
in the text area or the vector area, the pixels having color values
not corresponding to the index colors and the pixels contained in
the photograph area are contained in the bit map image data for
lossless compression.
[0132] Since the pixels contained in the index image data for
lossless compression have been extracted from the bit map image
data for lossless compression, when it is assumed that the pixel
values of the extracted pixels are 255 (white) for RGB, edge
components are generated at portions from which the pixels are
extracted. Such a bit map image data for lossless compression may
lead to reduction in compression efficiency, thereby being
undesirable in some cases. To prevent this, pixel values estimated
by pixel value estimation in the lossless compression method (the
JPEG-LS method in Embodiment 1) used in the second lossless
compression section 47 may be assigned to the pixels (the pixels
contained in the index image data for lossless compression)
extracted from the bit map image data for lossless compression.
[0133] The lossy compression section 45 performs the lossy
compression process according to the JPEG method for the bit map
image data for lossy compression obtained from the photograph image
segmentation section 43, and generates lossy compression image
data. The lossy compression section 45 may perform the lossy
compression process according to the JPEG 2000 method other than
the JPEG method, for example. However, the method is not limited to
these methods.
[0134] The first lossless compression section 46 regards the index
image data for lossless compression obtained from the index image
segmentation section 44 as a binary image of four bit planes, and
performs the binary image lossless compression process according to
the JPEG method for each bit plane. First, the first lossless
compression section 46 segments the index image data for lossless
compression, for each bit plane. Since the index image data for
lossless compression has 4 bits per pixel, the first lossless
compression section 46 segments the index image data for lossless
compression into four planes of 1-bit image data having the
respective bits (bit 0, bit 1, bit 2 and bit 3) of each pixel
value.
[0135] The first lossless compression section 46 performs the
lossless compression process according to the JBIG method for the
segmented four bit planes sequentially, and generates first
lossless compression image data. The first lossless compression
section 46 may perform the compression process according to the
binary image lossless compression methods, such as the MH method,
the MR method (ITU-T recommendation T.4), the MMR method (ITU-T
recommendation T.6) and the JBIG2 method recommended by ITU-T, in
addition to the JPEG method. However, the method is not limited to
these methods.
[0136] The second lossless compression section 47 performs the
lossless compression process according to the JPEG-LS method for
the bit map image data for lossless compression obtained from the
index image segmentation section 44, and generates second lossless
compression image data. The second lossless compression section 47
may perform the compression process according to the multilevel
image lossless compression methods, such as the Lossless JPEG
method and JPEG 2000 method, in addition to the JPEG-LS method.
However, the method is not limited to these methods.
[0137] When the data output from the RIP process section 2 has been
input to the image compression section 4 through the input selector
3, the identification information compression section 48 obtains
the pixel identification information data input from the input
selector 3. Since the pixel identification information data is
information indicating that each pixel belongs to which one of four
kinds of areas (a text area, a vector area, a photograph area and
an unclassified area), the pixel identification information data
has the same size as that of the continuous tone bit map image
data, and has 2 bits (4 values) per pixel.
[0138] The identification information compression section 48
regards the pixel identification information data as a binary image
of two bit planes, and performs the binary image lossless
compression process according to the MRR method for each bit plane.
First, the identification information compression section 48
segments the pixel identification information data for each bit
plane. Since the pixel identification information data has 2 bits
per pixel, the identification information compression section 48
segments the pixel identification information data into two planes
of 1-bit image data, that is, a bit plane having the MSB (most
significant bit) of each pixel and a bit plane having the LSB
(least significant bit) of each pixel.
[0139] The identification information compression section 48
performs the lossless compression process according to the MMR
method for the segmented two bit planes sequentially, and generates
pixel identification information compression data. The
identification information compression section 48 may perform the
compression process according to the binary image lossless
compression methods, such as the MH method, the MR method, the JBIG
method and the JBIG2 method, in addition to the MRR method.
However, the method is not limited to these methods.
[0140] The process performed by the image compression section 4 of
the image forming apparatus 100 according to Embodiment 1 will be
described below based on flowcharts. FIGS. 9 and 10 are flowcharts
showing the procedure of the image compression process performed by
the image compression section 4.
[0141] When the input of the continuous tone bit map image data
from the input selector 3 is started, the image compression section
4 performs step S1, step S2 and steps S3 to S7 in parallel with one
another. The image compression section 4 first stores the
continuous tone bit map image data input sequentially from the
input selector 3 in the buffer memory 41 (at step S1).
[0142] In addition, the image compression section 4 refers to the
pixel identification information data input sequentially from the
input selector 3, and generates block segmentation information
corresponding to each block in the continuous tone bit map image
data (at step S2). The block is a block having totally 64 pixels
comprising 8 pixels in the main scanning direction and 8 pixels in
the sub scanning direction, and the block segmentation information
is information indicating whether lossless compression or lossy
compression is performed for each block. When the pixel
identification information data has not been input from the input
selector 3, the image compression section 4 generates information
indicating that lossy compression is performed, as the block
segmentation information corresponding to all the blocks.
[0143] The image compression section 4 judges whether the
continuous tone bit map image data input from the input selector 3
is scanner image data read from a document by the image input
section 11 (at step S3). When the pixel identification information
data is not input from the input selector 3, the image compression
section 4 can judge that the input continuous tone bit map image
data is scanner image data. When the image compression section 4
judges that the continuous tone bit map image data is scanner image
data (YES at step S3), the image compression section 4 skips steps
S4 to S7. When the image compression section 4 judges that the
continuous tone bit map image data is not scanner image data (NO at
step S3), the image compression section 4 performs step S4 and
steps S5 to S7 in parallel with one another.
[0144] The image compression section 4 refers to the pixel
identification information data, counts the number of pixels
classified by each color value for the pixels in the text and
vector areas in the line data of 8 lines to be stored in the buffer
memory 41, and creates a color histogram (at step S4). On the other
hand, the image compression section 4 segments the pixel
identification information data into two bit planes (at step S5).
Then, the image compression section 4 performs the lossless
compression process according to the MMR method for each block
plane (at step S6), and generates compressed pixel identification
information data. The image compression section 4 judges whether
the compression process for the pixel identification information
corresponding to the line data of 8 lines has been completed (at
step S7). When the image compression section 4 judges that the
compression process has not been completed (NO at step S7), the
procedure returns to step S5, and steps S5 to S7 are repeated. When
the image compression section 4 judges that the compression process
has been completed (YES at step S7), the procedure advances to the
subsequent step.
[0145] After steps S1 to S7 have been completed, block segmentation
information, a color histogram and compressed pixel identification
information data corresponding to the line data of 8 lines stored
in the buffer memory 41 are generated, and the procedure to be
performed by the image compression section 4 advances to the
subsequent step. Note that if the image compression section 4 has
generated the block segmentation information and the color
histogram, the procedure may advance to the subsequent step even
when the compressed pixel identification information data is being
generated.
[0146] Next, the image compression section 4 performs steps S8 and
S9 and steps S10 to S18 in parallel with one another. The image
compression section 4 generates bit map image data for lossy
compression from the line data of 8 lines stored in the buffer
memory 41 based on the block segmentation information generated at
step S2 (at step S8). The image compression section 4 performs the
lossy compression process according to the JPEG method for the
generated bit map image data for lossy compression (at step S9),
thereby generating lossy compression image data.
[0147] The image compression section 4 judges whether the
continuous tone bit map image data input from the input selector 3
is the scanner image data read from the document by the image input
section 11 (at step S10). When the image compression section 4
judges that the continuous tone bit map image data is the scanner
image data (YES at step S10), the image compression section 4 skips
steps S11 to S18. When the image compression section 4 judges that
the continuous tone bit map image data is not the scanner image
data (NO at step S10), the image compression section 4 performs
descending sorting for the color histogram created at step S4 based
on the number of pixels (at step S11).
[0148] The image compression section 4 determines 15 colors as
index colors based on the color histogram having been subjected to
the sorting (at step S12). For example, the image compression
section 4 determines 15 colors as index colors in descending order
of the number of pixels, and assigns color codes 1 to 15 to the
color values of the determined 15 colors serving as index colors
sequentially. Next, the image compression section 4 performs steps
S13 to S16 and steps S17 and S18 in parallel with one another.
[0149] The image compression section 4 generates index image data
for lossless compression and bit map image data for lossless
compression from the line data of 8 lines stored in the buffer
memory 41, based on the block segmentation information generated at
step S2 and the index colors determined at step S12 (at steps S13
and S17). The image compression section 4 segments the generated
index image data for lossless compression into four bit planes (at
step S14). Then, the image compression section 4 performs the
lossless compression process according to the JBIG method for each
bit plane (at step S15), thereby generating first lossless
compression image data.
[0150] The image compression section 4 judges whether the
compression process for the index image data for lossless
compression generated at step S13 has been completed (at step S16).
When the image compression section 4 judges that the compression
process has not been completed (NO at step S16), the procedure
returns to step S14, and steps S14 to S16 are repeated, and when
the image compression section 4 judges that the compression process
has been completed (YES at step S16), the procedure advances to the
subsequent step.
[0151] Furthermore, the image compression section 4 performs the
lossless compression process according to the JPEG-LS method for
the bit map image data for lossless compression generated at step
S17 (at step S18), and generates second lossless compression image
data. After steps S8 to S18 have been completed, based on the line
data of 8 lines stored in the buffer memory 41 at step S1, lossy
compression image data, first lossless compression image data,
second lossless compression image data, index color information and
compressed pixel identification information data are generated.
[0152] The image compression section 4 judges whether the
above-mentioned process steps have been completed for all of the
continuous tone bit map image data input from the input selector 3
(at step S19). When the image compression section 4 judges that the
process steps for all of the continuous tone bit map image data
have not been completed (NO at step S19), the procedure to be
performed by the image compression section 4 returns to step S1,
and the above-mentioned process steps are repeated until the
process steps for all of the continuous tone bit map image data are
completed. When the image compression section 4 judges that the
process steps for all of the continuous tone bit map image data
have been completed (YES at step S19), the image compression
section 4 completes the above-mentioned process steps.
[0153] As described above, based on the pixel identification
information data shown in FIG. 7B, the image compression section 4
generates the bit map image data for lossy compression shown in
FIG. 8A, the index image data for lossless compression shown in
FIG. 8B and the bit map image data for lossless compression shown
in FIG. 8C from the continuous tone bit map image data shown in
FIG. 7A. Then, the image compression section 4 performs compression
process that differs according to the bit map image data for lossy
compression, the index image data for lossless compression and the
bit map image data, for lossless compression, for the bit map image
data for lossy compression, the index image data for lossless
compression and the bit map image data for lossless compression.
Since compression process suited for each kind of data is
performed, efficient compression process can be attained for the
entire image in consideration of the compression rate in each type
of compression process and the throughput and resources required
for the process.
[0154] Next, the image decompression section 6 will be described
below. The lossy compression image data, the first lossless
compression image data, the second lossless compression image data,
the index color information and the compressed pixel identification
information data generated by the image compression section 4 are
input from the storage section 5 to the image decompression section
6. When the image compression section 4 has compressed the scanner
image data read from the document by the image input section 11,
only the lossy compression image data is obtained as compressed
data. In this case, the image decompression section 6 obtains only
the lossy compression image data from the storage section 5.
[0155] Based on the input lossy compression image data, first
lossless compression image data, second lossless compression image
data, index color information and compressed pixel identification
information data, the image decompression section 6 generates
decompressed continuous tone bit map image data and pixel
identification information data. The image decompression section 6
outputs the decompressed continuous tone bit map image data and the
pixel identification information data that have been generated, to
the output selector 7 in the subsequent stage. The image
decompression section 6 according to Embodiment 1 performs
decompression process by using the line data of 8 lines in the main
scanning direction as one process unit, as in the case of the image
compression section 4. Furthermore, when the image decompression
section 6 has obtained only lossy compression image data, the image
decompression section 6 performs decompression process for the
lossy compression image data, and outputs the decompressed data to
the output selector 7 in the subsequent stage.
[0156] FIG. 11 is a block diagram showing the configuration of the
image decompression section 6. The image decompression section 6 is
equipped with an identification information decompression section
61, a lossy decompression section 62 (image decompression section),
a buffer memory 63, a first lossless decompression section 64
(image decompression section), a second lossless decompression
section 65 (image decompression section), an image composing
section 66, etc.
[0157] The identification information decompression section 61
obtains the compressed pixel identification information data from
the storage section 5. The identification information decompression
section 61 performs, for the compressed pixel identification
information data, decompression process according to the same
method as the method according to the compression process performed
by the identification information compression section 48 of the
image compression section 4. More specifically, the identification
information decompression section 61 performs the decompression
process according to the MMR method for the respective compressed
binary image data corresponding to two bit planes constituting the
compressed pixel identification information data. Furthermore, the
identification information decompression section 61 combines the
decompressed two bit planes, thereby sequentially restoring the
pixel identification information having 2 bits (4 values) per pixel
and generating pixel identification information data.
[0158] The lossy decompression section 62 performs lossy
decompression process according to the JPEG method for the lossy
compression image data obtained from the storage section 5 and
generates lossy decompression image data. Each piece of pixel data
of the lossy decompression image data is represented using three
color components of RGB, and each color component is represented by
256 gray scales that can be represented by 8 bits. Although the
lossy decompression image data is approximate to the bit map image
data for lossy compression before compression, since the
compression method is a lossy compression method, the lossy
decompression image data is not completely equal to the bit map
image data for lossy compression before compression. The lossy
decompression section 62 stores the generated lossy decompression
image data in the buffer memory 63 sequentially.
[0159] The buffer memory 63 temporarily stores the lossy
decompression image data having been decompressed by the lossy
decompression section 62. The buffer memory 63 outputs the lossy
decompression image data to the image composing section 66 together
with the index image data decompressed by the first lossless
decompression section 64, the bit map image data decompressed by
the second lossless decompression section 65 and the input index
color information. Although the lossy decompression section 62
performs decompression process according to the JPEG method in
block units, the first lossless decompression section 64 and the
second lossless decompression section 65 perform decompression
process in pixel units by scanning pixels in the main scanning
direction from the upper left of an image and by scanning the
adjacent line in the sub scanning direction after the scanning of
one line. Hence, the buffer memory 63 is used to output the lossy
decompression image data to the image composing section 66 in
synchronization with the output of the other decompressed data.
[0160] The first lossless decompression section 64 obtains the
first lossless compression image data from the storage section 5.
The first lossless decompression section 64 performs the
decompression process according to the JBIG method for the
compressed binary image data corresponding to four bit planes
constituting the first lossless compression image data.
Furthermore, the first lossless decompression section 64 combines
the decompressed four bit planes, thereby restoring the index image
data represented by the color codes having 4 bits (16 values) per
pixel.
[0161] The second lossless decompression section 65 performs the
lossy decompression process according to the JPEG-LS method for the
second lossless compression image data obtained from the storage
section 5, and generates bit map image data. Each piece of pixel
data of the decompressed bit map image data is represented using
three color components of RGB, and each color component is
represented by 256 gray scales that can be represented by 8
bits.
[0162] The index color information, the lossy decompression image
data, the index image data and the bit map image data are input to
the image composing section 66 in synchronization. The image
composing section 66 generates decompressed continuous tone bit map
image data based on the respective input data. The image composing
section 66 superimposes the lossy decompression image data, the bit
map image data and the index image data in this order, thereby
composing the respective data and generating the decompressed
continuous tone bit map image data.
[0163] More specifically, the image composing section 66 first
judges whether the pixel value (color code) of each pixel in the
index image data is value 0 indicating a transparent color. When
the pixel value is not the value indicating a transparent color,
that is, when the color of a pixel in the index image data is not a
transparent color, the image composing section 66 sets the color
value corresponding to the color code of each pixel and contained
in the index color information to the pixel value after the
composition. When the color is a transparent color, that is, when
the color of a pixel in the index image data is a transparent
color, the image composing section 66 judges whether the pixel
value of each pixel in the bit map image data corresponding to the
position of each pixel is value 255 (white) for RGB.
[0164] When the color is not white, that is, when the color of the
pixel in the bit map image data is not white, the image composing
section 66 sets the pixel value of each pixel in the bit map image
data to the pixel value obtained after the composition. When the
color is white, that is, when the color of the pixel in the bit map
image data is white, the image composing section 66 assumes that
the pixel is a transparent pixel, and sets the pixel value of each
pixel in the lossy decompression image data corresponding to the
position of each pixel to the pixel value after the composition. As
described above, in Embodiment 1, the highest priority is given to
the index image data and the next priority is given to the bit map
image data, and the image composing section 66 superimposes the
respective kinds of image data, thereby generating the decompressed
continuous tone bit map image data.
[0165] As described above, the index image data is superimposed on
the bit map image data for composition. Hence, among the pixels of
the bit map image data, pixels not coming out to the surface when
overwritten by the pixels of the index image data can be set to
appropriate pixel values to improve the compression ratio for the
bit map image data.
[0166] Although the decompressed continuous tone bit map image data
is approximate to the continuous tone bit map image data before
compression by the image compression section 4, when the
decompressed continuous tone bit map image data contains pixels
compressed by the lossy compression method and then decompressed
later, the decompressed continuous tone bit map image data is not
completely equal to the continuous tone bit map image data before
compression. When the decompressed continuous tone bit map image
data does not contain pixels compressed by the lossy compression
method and then decompressed later, the continuous tone bit map
image data before compression is equal to the decompressed
continuous tone bit map image data.
[0167] The process to be performed by the image decompression
section 6 of the image forming apparatus 100 according to
Embodiment 1 will be described below based on flowcharts. FIGS. 12
and 13 are flowcharts showing the procedure of the image
decompression process by the image decompression section 6.
[0168] On starting the reading of the data stored in the storage
section 5, the image decompression section 6 performs steps S21 and
S22 and steps S23 to S26 in parallel with one another. The image
decompression section 6 first performs the lossy decompression
process according to the JPEG method for the lossy compression
image data read from the storage section 5 (at step S21), thereby
generating lossy decompression image data. The image decompression
section 6 stores the generated lossy decompression image data into
the buffer memory 63 sequentially (at step S22).
[0169] On the other hand, the image decompression section 6 judges
whether the compressed data to be processed is data obtained by
compressing the scanner image data read from the document by the
image input section 11 (at step S23). When the compressed data to
be processed is the scanner image data, the image decompression
section 6 obtains only the lossy compression image data from the
storage section 5. Hence, when the image decompression section 6
has obtained only the lossy compression image data from the storage
section 5, the image decompression section 6 judges that the
compressed data to be processed is the scanner image data.
[0170] When the image decompression section 6 judges that the
compressed data to be processed is the scanner image data (YES at
step S23), the image decompression section 6 skips steps S24 to
S26. When the image decompression section 6 judges that the
compressed data to be processed is not the scanner image data (NO
at step S23), the image decompression section 6 performs the
lossless decompression process according to the MMR method for the
respective two bit planes constituting the pixel identification
information data (at step S24). The image decompression section 6
judges whether the decompression process for the two bit planes has
been completed (at step S25). When the image decompression section
6 judges that the decompression process has not been completed (NO
at step S25), the procedure returns to step S24, and the
decompression process continues.
[0171] When the image decompression section 6 judges that the
decompression process has been completed (YES at step S25), the
image decompression section 6 combines the decompressed two bit
planes (at step S26) to restore the pixel identification
information data. After steps S21 to S26 have been completed, the
decompression process for the pixel identification information data
is completed, and lossy decompression image data for the line data
of 8 lines is stored in the buffer memory 63.
[0172] Next, the image decompression section 6 judges whether the
compressed data to be processed is the scanner image data (at step
S27). When the image decompression section 6 judges that the
compressed data is the scanner image data (YES at step S27), the
procedure advances to step S36. When the image decompression
section 6 judges that the compressed data to be processed is not
the scanner image data (NO at step S27), the image decompression
section 6 performs steps S28 to S30 and step S31 in parallel with
one another.
[0173] The image decompression section 6 performs the lossless
decompression process according to the JBIG method for the
respective four bit planes constituting the first lossless
compression image data (at step S28). The image decompression
section 6 judges whether the decompression process for the four bit
planes has been completed (at step S29). When the image
decompression section 6 judges that the decompression process has
not been completed (NO at step S29), the procedure returns to step
S28, the decompression process continues. When the image
decompression section 6 judges that the decompression process has
been completed (YES at step S29), the image decompression section 6
combines the decompressed four bit planes (at step S30), thereby
generating index image data.
[0174] The image decompression section 6 performs the lossless
decompression process according to the JPEG-LS method for the
second lossless compression image data (at step S31), thereby
generating bit map image data. After steps S28 to S31 have been
completed, the image decompression section 6 completes the
decompression process for all the compressed data based on the line
data of 8 lines.
[0175] The image decompression section 6 judges whether the color
code (pixel value) of each pixel is value 0 indicating a
transparent color in the index image data generated by combining
the bit planes at step S30 (at step S32). When the image
decompression section 6 judges that the color code does not
indicate a transparent color (NO at step S32), the image
decompression section 6 sets the color value indicating the color
code of each pixel, not indicating a transparent color, to a pixel
value after the composition (at step S33). When the image
decompression section 6 judges that the color code indicates a
transparent color (YES at step S32), the image decompression
section 6 judges whether the pixel value of each pixel in the bit
map image data corresponding to the position of each pixel
contained in the index image data and having a color code
indicating a transparent color has a white value (value 255 for
RGB) (at step S34).
[0176] When the image decompressing section 6 judges that the pixel
value is not the white value (NO at step S34), the image
decompression section 6 sets the pixel value of each pixel of the
bit map image data to a pixel value after the composition (at step
S35). When the image decompressing section 6 judges that the pixel
value is the white value (YES at step S34), the image decompression
section 6 sets the pixel value of each pixel in the lossy
decompression image data corresponding to the position of each
pixel contained in the bit map image data and having a color code
indicating white, to a pixel value after the composition (at step
S36). On the other hand, when the image decompression section 6
judges that the compressed data to be processed is the scanner
image data (YES at step S27), the image decompression section 6
sets the pixel value of each pixel in the lossy decompression image
data to a pixel value after the composition (at step S36).
[0177] The image decompression section 6 judges whether the
above-mentioned steps S27 to S36 have been completed for all the
pixels in the line data of 8 lines (at step S37). When the image
decompression section 6 judges that the steps for all the pixels
have not been completed (NO at step S37), the image decompression
section 6 returns the procedure to step S27, and the
above-mentioned steps are repeated. As a result, decompressed
continuous tone bit map image data of the line data of 8 lines is
generated.
[0178] When the image decompression section 6 judges that the steps
for all the pixels have been completed (YES at step S37), the image
decompression section 6 judges whether the above-mentioned steps
have been completed for all the compressed data read from the
storage section 5 (at step S38). When the image decompression
section 6 judges that the steps for all the compressed data have
not been completed (NO at step S38), the image decompression
section 6 returns the procedure to step S21, and the
above-mentioned steps are repeated until the steps for all the
compressed data are completed. As a result, decompressed continuous
tone bit map image data for one plate is generated. When the image
decompression section 6 judges that the steps for all the
compressed data have been completed (YES at step S 38), the image
decompression section 6 completes the above-mentioned steps.
[0179] At step S32 for the above-mentioned decompression process,
the image decompression section 6 judges whether the color code of
each pixel is value 0 indicating a transparent color in the index
image data. However, when the index color information has color
codes of less than 16 colors, the image decompression section 6 may
judge that only less than 16 colors are present in the lossless
compression area in the continuous tone bit map image data
including the bit map image data, and steps S34 and S35 may be
skipped.
[0180] FIG. 14 is a table showing the effects obtained when the
present invention is applied. FIG. 14 shows the results obtained by
evaluating the effects of improvement in compression rate in
comparison with the related art in the case that the art described
in Embodiment 1 is applied to general print output image
information at which the art of the present invention is aimed. In
addition, FIG. 14 shows examples of the effects obtained by
applying the present invention to three test sets 1, 2 and 3 used
as sample sets. Image samples contained in the respective test sets
1, 2 and 3 are different from one another. The test set 1 contains
12 image samples, the test set 2 contains 20 image samples, and the
test set 3 contains 12 image samples.
[0181] As the evaluation results, FIG. 14 shows the total number of
pixels A constituting all the image samples contained in each of
the test sets 1, 2 and 3, and also shows lossless compression
target ratio B (%) indicating the ratio of the number of pixels
targeted for lossless compression with respect to the total number
of pixels A. Furthermore, FIG. 14 shows index image covering ratio
C (%) indicating the ratio of the number of pixels capable of being
contained actually in index image data represented by 4 bits per
pixel with respect to the total number of pixels A. Moreover, for
all the image samples contained in the test sets 1, 2 and 3 in the
case that the art according to Embodiment 1 is applied, FIG. 14
shows the total number of bytes of lossless compression data D
obtained by adding the number of bytes of the first lossless
compression image data compressed by the first lossless compression
section 46, the number of bytes of the second lossless compression
image data compressed by the second lossless compression section 47
and the number of the bytes of the index color information.
[0182] Furthermore, in comparison with the related art, FIG. 14
shows the results obtained when all the areas to be subjected to
lossless compression in the continuous tone bit map image data are
contained in the bit map image data for lossless compression
without generating index image data for lossless compression at the
index image segmentation section 44. More specifically, when all
the areas to be subjected to lossless compression in the continuous
tone bit map image data are subjected to lossless compression by
the second lossless compression section 47, for all the image
samples respectively contained in the test sets 1, 2 and 3, FIG. 14
shows the total number E of bytes of JPEG-LS compressed data
obtained by adding the numbers of bytes of the second lossless
compression image data compressed by the second lossless
compression section 47. Still further, FIG. 14 shows lossy
compression progress rate F that is obtained by dividing E by D
shown in FIG. 14.
[0183] As shown in FIG. 14, by using the art described in
Embodiment 1 for general print output image information, most
information contained in areas to be subjected to lossless
compression in the continuous tone bit map image data can be
contained in an index image having a limited amount of information.
This is because the number of colors used in the text and vector
areas contained in the general print output image information is
usually not large, as shown in FIG. 6. Hence, the compression rate
can be improved at very high efficiency in comparison with the
related art as indicated by the lossy compression progress rate F
shown in FIG. 14, by allowing numerous areas to be subjected to
lossless compression in the continuous tone bit map image data to
be contained in the index image.
[0184] On the other hand, even when the number of color codes (the
number of colors) is set to 1024 as shown in FIG. 6, the continuous
tone bit map image data cannot be covered 100%. Generally speaking,
although PDL data is created using a software application in an
external PC or the like, the user who creates the PDL data usually
tends not to use numerous colors for texts and vectors (graphics).
However, there are various cases; a case in which tone images
wherein gray scale and hue changes stepwise are used; a case in
which a different color is used for each dot; or a case in which a
multicolor pattern typified by a color evaluation chart, such as
the Macbeth chart, is used. Since numerous colors are used locally
in such a case, the number of gray scales or the number of colors
of the input image cannot be limited in the image compression
process for the text and vector areas to be subjected to lossless
compression.
[0185] Hence, in Embodiment 1, in the continuous tone bit map image
data, a photograph area which contains numerous colors and whose
compression rate is hardly expected to be improved by lossless
compression is first segmented in data for lossy compression. Then,
the remaining text and vector areas are used as areas to be
subjected to lossless compression. Color histograms are created for
these areas, and pixels (areas) in which colors having high
frequency of occurrence are further extracted and the pixel values
thereof are replaced with color codes, whereby lossless compression
index image data is obtained.
[0186] In addition, the compression process according to the JBIG
method is used for the lossless compression method for index image
data containing pixels in which colors having high frequency of
occurrence are used. The compression process according to the JBIG
method is performed by using an arithmetic code compression method
referred to as QMx coder as entropy coding together with numerous
context classifications, and the compression rate is expected to be
improved so as to be relatively high. On the other hand, a
relatively large amount of process resources is required. However,
the required amount of the process resources can be reduced by
decreasing the number of bits per pixel by performing the
compression process.
[0187] On the other hand, the lossless compression process
according to the JPEG-LS method and capable of being attained
without requiring relatively large amount of process resources is
performed for image data in areas not contained in index image data
among the text and vector areas to be subjected to lossless
compression. Since main areas in which index colors are used, more
specifically, areas containing edge components and the like and
being hardly expected to have a high compression rate, have already
been taken out to the index image, even if the compression process
according to the JPEG-LS method that is hardly expected to
relatively improve a compression ratio is performed for the
remaining areas, the compression rate can be improved.
[0188] Furthermore, resource reduction and improvement in
compression rate are attained for the photograph area by using the
JPEG method serving as a lossy compression method suited for the
photograph area. In the case of an image whose area is essentially
and wholly constituted by a photograph area as in the case of an
image obtained by a scanner or the like, the same image compression
section 4 and image decompression section 6 can be used by using
only the lossy compression process method.
[0189] As described above, a process method that utilizing the
characteristics of a printing image having numerous objects, such
as a text, a graphic and a photograph, is used in Embodiment 1.
Hence, the amount of resources required for each type of process
can be balanced with the improvement in compression rate while the
reversibility for an area to be subjected to lossless compression
is maintained.
Embodiment 2
[0190] In the above-mentioned Embodiment 1, an area to be subjected
to lossless compression in continuous tone bit map image data to be
compressed is segmented into two areas (index image data for
lossless compression and bit map image data for lossless
compression) by the image compression section 4, and lossless
compression methods different from each other are used for the two
areas. Hence, the amount of resources required for each type of
process can be balanced with the improvement in compression rate in
each type of process.
[0191] In addition to the above-mentioned configuration, it may be
possible to have a configuration in which an area to be subjected
to lossless compression in continuous tone bit map image data is
broken down more finely, for example, the area to be subjected to
lossless compression is segmented into three kinds of image data
based on the frequency of occurrence of colors to be used. In this
case, for example, as a first segmentation image, pixels having
three colors used at the highest frequency (when the total number
of colors is more than four; in this case, one color is a
transparent color) or having four colors used at the highest
frequency (when the total number of colors is less than four) are
extracted as index colors appearing most frequently. Furthermore, 2
bits per pixel are assigned to the extracted pixels having
respective colors, each assigned bit is regarded as a pixel
constituting a bit plane, and each of two bit planes is compressed
using the JBIG method.
[0192] Furthermore, as a second segmentation image, pixels having
255 colors used at the next highest frequency (when the total
number of colors is more than 259; in this case, one color is a
transparent color) or having 256 colors used at the next highest
frequency (when the total number of colors is less than 259) are
extracted as index colors appearing frequently. Moreover, 8 bits
per pixel are assigned to the extracted pixels having respective
colors, and compression is performed using the DEFLATE method while
8 bits are used as one pixel. Still further, as a third
segmentation image, the pixels having the remaining colors are
compressed using the JPEG-LS method. However, as the compression
process for the second segmentation image, methods other than the
DEFLATE method may also be used. An appropriate compression method
should only be selected based on the amount resources required for
the process and the compression rate in the process.
[0193] As described above, an area to be subjected to lossless
compression in the continuous tone bit map image data may be
segmented into three or more images, and compression methods
different from one another may be used. When the area is segmented
into three or more images, judgment priority at the time of the
segmentation may be changed using information other than the
frequency of occurrence of color, such as a flag indicating whether
each color is used in a text area. When the area to be subjected to
lossless compression is segmented into three or more images, as
shown in FIG. 6, a covering ratio more than 80% of the area to be
subjected to lossless compression can be obtained only by the first
segmentation image, and most of the area to be subjected to
lossless compression can be covered by the first and second
segmentation images. Hence, in Embodiment 2, the compression rate
can be balanced with the required amount of resources more
finely.
[0194] In Embodiments 1 and 2 described above, a color histogram is
created for the line data of 8 lines, and the colors (index colors)
of pixels contained in index image data for lossless compression
are determined based on the created color histogram. However, the
method for determining the index colors is not limited to this kind
of method.
[0195] For example, it may be possible that the color information
of the colors used for the printer driver of the external PC has
been transmitted to the image forming apparatus 100 in advance and
that the image forming apparatus 100 determines the colors
indicated by the received color information as index colors.
Furthermore, it may also be possible that the color information of
the colors used in PDL data is inserted into the PDL data and the
printer driver of the external PC transmits the color information
together with the PDL data to the image forming apparatus 100. In
this case, the RIP process section 2 of the image process section
10 should only transmit the received color information to the image
compression section 4 directly, and the image compression section 4
should only determine index colors based on the obtained color
information.
[0196] In addition to the configurations according to Embodiments 1
and 2 described above, another configuration may be used
additionally in which lossless compression process is performed for
the entire continuous tone bit map image data to be compressed
according to the designation from the user. In other words, even
image data containing areas neither a text area nor a vector area
is classified into index image data for lossless compression and
bit map image data for lossless compression, and lossless
compression is performed for each of the two kinds of image data.
Hence, the reversibility of the entire image data can be obtained
securely, and the degradation in image quality before and after
compression process can be prevented. However, since the colors
used for the pixels in a photograph area are not used for the
creation of a color histogram, the photograph area is classified as
the bit map image data for lossless compression. Hence, the amount
of data classified as the index image data for lossless compression
can be prevented from increasing, whereby the increase of the
process load due to use of the first lossless compression section
46 can be reduced.
[0197] In addition, the configuration according to Embodiment 2 in
which image data is classified into three kinds of data, that is,
bit map image data for lossy compression, index image data for
lossless compression and bit map image data for lossless
compression, or the configuration according to Embodiment 1 in
which image data is classified into two kinds of data, that is,
index image data for lossless compression and bit map image data
for lossless compression may be selected by switching according to
the designation from the user. For example, when the user issues a
print request from the external PC to the image forming apparatus
100 and the user selects "lossy compression prohibition" or "high
quality" through the printer driver, image data may be classified
into two kinds of data. On the other hand, when the user issues a
print request and the user selects "lossy compression execution" or
"ordinary compression" through the printer driver, image data may
be classified into three kinds of data.
[0198] Furthermore, it may be possible that the configuration in
which image data is classified into two kinds of data or the
configuration in which image data is classified into three kinds of
data has been registered in the image forming apparatus 100 for
each user and that the configuration corresponding to the user who
requests printing is used for image compression. Moreover, it may
also be possible that the configuration in which image data is
classified into two kinds of data or the configuration in which
image data is classified into three kinds of data has been
registered in the image forming apparatus 100 for each external PC
and that the configuration corresponding to the external PC (PC
from which the PDL data was sent) which requests printing is used
for image compression. Still further, the administrator of each
image forming apparatus 100 may set whether image data is
classified into two or three kinds of data for each image forming
apparatus 100.
Embodiment 3
[0199] In Embodiments 1 and 2 Described Above, a Configuration is
described in which the image compressing apparatus and the image
decompressing apparatus according to the present invention are
mounted in the same image forming apparatus 100. In Embodiment 3, a
configuration is described in which the image compressing apparatus
according to the present invention is mounted in a PC and the image
decompressing apparatus according to the present invention is
mounted in the image forming apparatus.
[0200] FIG. 15 is a block diagram showing the configuration of the
main section of a PC according to Embodiment 3. A PC 200 according
to Embodiment 3 is equipped with a PDL data generation section 201,
an RIP process section 202, an image compression section 203, a
storage section 204, an external interface section 205, etc. The
RIP process section 202, the image compression section 203, the
storage section 204 and the external interface section 205 provided
for the PC 200 according to Embodiment 3 have the same
configurations as those of the RIP process section 2, the image
compression section 4, the storage section 5 and the external
interface section 13 provided for the image process section 10
according to Embodiment 1 described above, respectively, and
perform similar process.
[0201] The PDL data generation section 201 is a printer driver, for
example, and converts created data into data written in a page
description language (PDL language), thereby generating PDL data.
Data to be converted into the PDL data is created, for example,
using a text editing function attained by executing application
software, such as text editing software or image editing software,
stored in advance in the hard disk of the PC 200 by the CPU of the
PC 200.
[0202] The PDL data generated by the PDL data generation section
201 is subjected to RIP process at the RIP process section 202 and
converted into continuous tone bit map image data. The RIP process
section 202 is used to generate the continuous tone bit map image
data and pixel identification information data. The two kinds of
data generated by the RIP process section 202 are subjected to
image compression by the process described in Embodiment 1 at the
image compression section 203, and stored in the storage section
204. The compression data stored in the storage section 204 is
transmitted from the external interface section 205 to the image
forming apparatus 100 via a network or the like at predetermined
timing.
[0203] FIG. 16 is a block diagram showing the configuration of an
image forming apparatus 100 according to Embodiment 3. The image
forming apparatus 100 (for example, a digital color printer)
according to Embodiment 3 is equipped with an image process section
10, an image output section 12, an external interface section 13,
an operation panel 14, etc. Furthermore, the image process section
10 according to Embodiment 3 is equipped with a storage section 5,
an image decompression section 6, an output selector 7, a print
image process section 8, a transmission image process section 9,
etc. The image forming apparatus 100 according to Embodiment 3 has
a configuration similar to that of the image forming apparatus 100
according to Embodiment 1 described above, and performs similar
process. However, the image forming apparatus 100 according to
Embodiment 3 is not equipped with the image input section 11
provided for the image forming apparatus 100 according to
Embodiment 1 described above, and the image process section 10
according to Embodiment 3 is not equipped with the scanner image
process section 1, the RIP process section 2, the input selector 3
and the image compression section 4 provided for the image process
section 10 according to Embodiment 1 described above.
[0204] When the image forming apparatus 100 according to Embodiment
3 receives compressed data from the external PC 200 via the
external interface section 13, the image forming apparatus 100
stores the received compressed data in the storage section 5. The
compressed data stored in the storage section 5 is read out by the
image decompression section 6 in the subsequent stage, and
decompressed using process similar to that described in Embodiment
1.
[0205] As described above, when data subjected to image compression
by the PC 200 is subjected to image decompression by the image
forming apparatus 100, since the amount of data to be transmitted
from the PC 200 to the image forming apparatus 100 can be reduced,
the communication load on the network can be reduced. In addition,
the process load on the image forming apparatus 100 can also be
reduced by distributing compression process and decompression
process to the PC 200 and the image forming apparatus 100.
[0206] On the other hand, it is conceivable that the compressed
data stored in the storage section 5 of the image forming apparatus
100 is transmitted to the external PC 200 via the external
interface section 13 and reused by the PC 200. Hence, the PC 200
may be further equipped with the function of the image
decompression section 6 of the image forming apparatus 100 in
addition to the configuration shown in FIG. 15. Furthermore, it is
conceivable to use a configuration in which the PC 200 does not
perform compression process but performs only decompression process
for the compressed data received from the image forming apparatus
100. In this case, it may be possible that the PC 200 is equipped
with neither the RIP process section 202 nor the image compression
section 203 shown in FIG. 15, but is equipped with only the
function of the image decompression section 6 of the image forming
apparatus 100.
[0207] In the respective embodiments described above, the
components including the image compression sections 4 and 203 and
the image decompression section 6 provided for the image forming
apparatus 100 or the PC 200 may be embodied by hardware logic or
may be embodied by software using a processor, such as a CPU or an
MPU. In other words, the image forming apparatus 100 is equipped
with a CPU for executing the instructions of control programs for
accomplishing various functions, a ROM for storing the control
programs, a RAM (random access memory) to which various kinds of
control programs are loaded, a storage unit (recording medium),
such as a memory, for storing various kinds of control programs and
various kinds data, etc.
[0208] The object of the present invention is also attained by
supplying, to the image forming apparatus 100, a recording medium
on which the program codes (executable programs, intermediate code
programs and source programs) of control programs for accomplishing
various functions are recorded and which can be read by computers,
and by allowing the image forming apparatus 100 (CPU pr MPU) to
read and execute the programs codes recorded on the recording
medium. In this case, the recording medium on which the computer
programs for accomplishing the image compression method or the
image decompression method according to the present invention can
be provided so as to be portable.
[0209] The recording medium may be a memory, not shown, intended
for microcomputer process, for example, a program medium, such as a
ROM, or may also be a program medium that can be read when the
recording medium is inserted into a program reader serving as an
external storage unit, not shown.
[0210] In any case, it may be possible that the stored program
codes are accessed by a microprocessor and executed, or it may be
possible that the program codes are read out, and the read-out
program codes are downloaded in the program storage area, not
shown, of the microcomputer, and then the program codes are
executed. In this case, it is assumed that the computer program for
the downloading has been stored in the main apparatus in
advance.
[0211] The above-mentioned program medium is a recording medium
that can be separated from the main apparatus and may be a medium
capable of fixedly storing computer programs. Examples of the
medium may include a tape type, such as magnetic tape and cassette
tape; a disc type, for example, a magnetic disk, such as a flexible
disk and a hard disk, and an optical disc, such as CD-ROM, MO, MD,
DVD and CD-R; a card type, such as an IC card (including a memory
card) and an optical card; or a semiconductor memory, such as mask
ROM, EPROM (erasable programmable read only memory), EEPROM
(electrically erasable programmable read only memory) and flash
ROM.
[0212] In addition, in the respective embodiments described above,
since the image forming apparatus 100 and the PC 200 are configured
so as to be connectable to a communication network including the
Internet, the recording medium may be a medium for flexibly holding
the program codes by downloading the program codes via the
communication network. When the program codes are downloaded from
the communication network, the program for performing the
downloading may have been stored in the main apparatus in advance
or installed from another recording medium.
[0213] Furthermore, the communication network is not particularly
limited. For example, it is possible to use the Internet, intranet,
extranet, LAN, ISDN, VAN, CATV communication network, virtual
private network, telephone network, mobile communication network,
satellite communication network, etc. Moreover, a transmission
medium constituting the communication network is not particularly
limited. For example, it is possible to use wired communication
lines, such as IEEE 1394, USB, power line transmission, cable TV
line, telephone line and ADSL line, or wireless communication, for
example, communication using infrared rays for IrDA and remote
control, Bluetooth (registered trademark), 802.11 wireless
communication, HDR, mobile telephone network, satellite circuit and
terrestrial digital network. The present invention can also be
accomplished by computer data signals embedded in a carrier wave in
which the above-mentioned program codes are embodied through
electronic transmission.
[0214] Still further, the components provided for the image forming
apparatus 100 are not limited to have configurations embodied by
hardware logic or by software. For example, it may be possible that
part of the components of the image forming apparatus 100 is
embodied by hardware, and that the remaining parts thereof and the
control of the hardware are embodied by software.
[0215] In the respective embodiments described above, as the image
input section 11, a flat bed scanner, a film scanner, a digital
camera or a mobile telephone is used, for example. Furthermore, as
the image output section 12, not only a printer, but also an image
display apparatus, such as a CRT display or a liquid crystal
display, is used.
[0216] Although the preferred embodiments according to the present
invention have been described specifically, the configurations,
operations, etc. thereof can be changed as necessary, and the
embodiments are not limited to the above-mentioned embodiments.
[0217] As this description may be embodied in several forms without
departing from the spirit of essential characteristics thereof, the
present embodiment is therefore illustrative and not restrictive,
since the scope is defined by the appended claims rather than by
the description preceding them, and all changes that fall within
metes and bounds of the claims, or equivalence of such metes and
bounds thereof are therefore intended to be embraced by the
claims.
* * * * *